THE ORGANIZATIONAL RESPONSE TO THE WORK-FAMILY NEEDS OF THE
MODERN WORKFORCE
By
Matthew Michael Piszczek

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Industrial Relations & Human Resources – Doctor of Philosophy
2014

ABSTRACT
THE ORGANIZATIONAL RESPONSE TO THE WORK-FAMILY NEEDS OF THE
MODERN WORKFORCE
By
Matthew Michael Piszczek
As the technological landscape and labor market demographics change, organizations are
finding themselves faced with a need to act upon employees’ increasing demand to better
manage their work and family roles. Over the past thirty years, a vast repository of research has
sought to identify what organizations do to meet this challenge and how formal and informal
organizational policies and practices affect employees as well as the organizations that offer
them. Despite the rise in popularity of work-family policies and practices as a research topic
across academic disciplines, the focus has been on how organizational practices affect
individuals. However, this research does not provide a sufficiently nuanced account of the
reasoning behind and implementation of these practices and policies. In particular, researchers
lack theory and schematic frameworks at the organizational level to understand how a particular
work-family policy or practice is chosen or implemented, and how it fits within the broader goals
of the organization.
In light of this, this dissertation provides two primary contributions to the organizationalwork family literature. First, it includes two empirical studies which focus on modern
organizational work-family issues that are relatively poorly understood in current research. One
study demonstrates the importance of organizational expectations surrounding the use of
informational and communicative technology after normal work hours. This study finds that
organizations pressuring employees to use information and communicative technology to work

outside of normal work hours risk employee emotional exhaustion, but that individual
differences in the preferences for work-family role management play a part in determining the
severity of this risk. The second study discusses the importance of public policy in
contextualizing organizational work-family practices by identifying Germany as a country with
an under-supplied public daycare system. The study shows that in an environment where
childcare is in high demand, organizations are able to step in with childcare practices that
mutually benefit employees and their organizations, and that those organizations with high levels
of female turnover are most likely to adopt these practices.
Second, I use these two studies alongside other empirical and theoretical evidence to
build a new theoretical framework for the adoption and implementation of work-family policies
and practices. Previous frameworks have not been updated to include the wealth of research
around organizational policies and practices and their relationship with employee and
organizational outcomes, and do not allow for variation in the adoption and implementation of
practices and policies. In particular, existing frameworks assume a generalized organizational
response and cannot explain the complex systems of work-family policies and practice observed
in actual organizations. Building an integrated organizational work-family response framework
provides some much-needed organization-level theory in the work-family domain. I discuss the
weaknesses of existing frameworks as well as the benefits of an integrated framework for
research and practice.

DEDICATION
I dedicate this work to my family with deep gratitude for their unwavering and unconditional
support and encouragement. I could not have succeeded without them.

iv

ACKNOWLEDGMENTS

My thanks go out to my dissertation committee. Dr. Ellen Kossek and Dr. Ann Marie
Ryan provided valuable insights into the work-family literature and various aspects of scale
structure and study design. Dr. Dale Belman provided guidance on econometric analyses. I am
especially thankful to my adviser and dissertation chair, Dr. Peter Berg, for his involvement and
advice from start to finish in all aspects of this project as well as his consistent and constructive
feedback and general support throughout the dissertation process. Most importantly, I am
thankful to Dr. Berg for teaching me the meaning of scholarship. Finally, I would also like to
thank Dr. Bill Cooke and Kristi White for their help in sending out the survey used in one of the
studies, and the many School of Human Resources & Labor Relations faculty members who
provided insightful feedback and suggestions during various internal presentations of this work
as it progressed.

v

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................................... viii
LIST OF FIGURES ........................................................................................................................ x
1

Work, Family, and the Organization....................................................................................... 1
1.1 Introduction ...................................................................................................................... 1
1.2 Three Organizational Work-Family Response Frameworks ............................................ 5
1.2.1 Development ............................................................................................................. 5
1.2.2 The Institutional Pressure Framework ...................................................................... 7
1.2.3 The Rational Choice Perspective .............................................................................. 9
1.2.4 The Agency Perspective ......................................................................................... 11
1.3 Strengths and Weaknesses of the Organizational Work-Family Response Frameworks
………………………………………………………………………………………….13
1.4 Dissertation Structure ..................................................................................................... 19
1.5 Conclusion...................................................................................................................... 22

2

Two Faces of Technology: Organizational Expectations at the Work-Family Interface ..... 24
2.1 Introduction .................................................................................................................... 24
2.2 Technology at the Work-Family Interface ..................................................................... 24
2.3 The Two Faces of Technology ....................................................................................... 30
2.4 Hypothesis Development ............................................................................................... 36
2.5 Methods .......................................................................................................................... 44
2.5.1 Measures ................................................................................................................. 44
2.5.2 Pilot Survey Analysis .............................................................................................. 47
2.5.2.1 Scale Reliability……………………………………………………………….48
2.5.2.2 ICT Use Scale Construction…………………………………………………...48
2.5.3 Full Survey Analysis ............................................................................................... 50
2.5.3.1 Sample and Design..…………………………………………………………..50
2.5.3.2 Analytic Techniques…………………………………………………………..55
2.6 Results ............................................................................................................................ 59
2.7 Discussion ...................................................................................................................... 68
2.8 Limitations and Conclusions .......................................................................................... 76

3

Filling the Childcare Gap: Work-Family Practices and Turnover in Germany .................... 79
3.1 Introduction .................................................................................................................... 79
3.2 Germany and the Business Case for Childcare .............................................................. 80
3.3 Work-Family Practices and Strategic Human Resource Management .......................... 87
3.4 Conflicting Effects in Strategic Human Resource Management Theory ....................... 93
3.5 WFP Effectiveness and Gender...................................................................................... 98
3.6 Methods .......................................................................................................................... 99
3.6.1 Sample..................................................................................................................... 99
3.6.2 Variables ............................................................................................................... 100
vi

3.6.3 Analysis................................................................................................................. 101
3.7 Results .......................................................................................................................... 103
3.7.1 Variance Components Analysis ............................................................................ 103
3.7.2 Descriptive Statistics ............................................................................................. 104
3.7.3 Regression Analyses ............................................................................................. 107
3.8 Discussion .................................................................................................................... 114
3.9 Limitations and Conclusions ........................................................................................ 118
4

An Integrated Organizational Work-Family Response Framework ................................... 120
4.1 Introduction .................................................................................................................. 120
4.2 Chapters 2 and 3 and the Organizational Work-family Response ............................... 121
4.3 Benefits of an Updated Framework ............................................................................. 126
4.4 An Integrated Organizational Work-family Response Framework ............................. 133
4.5 Using the New Work-Family Response Framework ................................................... 146
4.6 Limitations and Future Research.................................................................................. 150
4.7 Final Conclusions ......................................................................................................... 151

APPENDICES ............................................................................................................................ 154
Appendix A: Work-family Technology Survey and Instructions ....................................155
Appendix B: Supplemental Analyses of German Childcare Practices ............................161
REFERENCES ........................................................................................................................... 165

vii

LIST OF TABLES

Table 1.1. Goodstein’s (1994) framework for strategic responses to institutional pressures ......... 9
Table 1.2: Kirchmeyer’s (1995) work-family response archetypes.............................................. 12
Table 2.1: Number of scale items and alpha coefficients for the pilot study................................ 49
Table 2.2: Work-family and family-work technology use factor structure matrix. ...................... 52
Table 2.3: Retained work-family and family-work technology use items.................................... 53
Table 2.4: Measure intercorrelations, alphas, means, and standard deviations ............................ 63
Table 2.5: Regression models ....................................................................................................... 65
Table 2.6: Simple slope coefficients and statistical significance for Figure 2.3........................... 67
Table 2.7: Simple slope coefficients and statistical significance for Figure 2.4........................... 68
Table 2.8: Bias-corrected bootstrap test of indirect effects of interaction between work-family
separation preferences and after-hours electronic communications expectations on
emotional exhaustion, through work-family conflict.................................................. 70
Table 2.9: Post-hoc regression analysis with three-way interaction ............................................. 73
Table 2.10: Bonferroni-corrected simple slope comparisons for three-way interaction of workfamily technology use, work-family separation preferences, and after-hours electronic
communications expectations on emotional exhaustion ............................................. 74
Table 3.1: One-way ANOVA results on dependent variables demonstrating within-and-between
establishment variance components .......................................................................... 104
Table 3.2: Variable descriptive statistics with between and within standard deviations ............ 105
Table 3.3: Frequency of establishment changes in childcare practices ...................................... 106
viii

Table 3.4: Breakdown of sample and childcare offerings by industry ....................................... 107
Table 3.5: Between-establishment regressions ........................................................................... 110
Table 3.6: Within-establishment regression models with establishment fixed effects ............... 111
Table 3.7: Combined linear model.............................................................................................. 112
Table 3.8: Comparison of undstandardized regression coefficients for male and female turnover
................................................................................................................................... 113
Table B.1: Within-establishment poisson regression models with establishment fixed effects..162
Table B.2: Between-establishment poisson regression models………………………………...163
Table B.3: Within- and between-establishment combined poisson regression model…………164

ix

LIST OF FIGURES

Figure 2.1: Theoretical model with hypotheses ............................................................................ 37
Figure 2.2: Scree plot for work-family and family-work technology use items. .......................... 51
Figure 2.3: Simple slopes of the after-hours electronic communications expectations and workfamily ICT use interaction on emotional exhaustion .................................................. 67
Figure 2.4: Simple slopes of the after-hours electronic communications expectations and workfamily segmentation preferences interaction on work-family conflict ....................... 67
Figure 2.5: Empirically supported model ..................................................................................... 70
Figure 2.6: Three-way interaction of work-family technology use, work-family separation
preferences, and after-hours electronic communications expectations on emotional
exhaustion ................................................................................................................... 74
Figure 4.1: An integrated model of the organizational work-family response ........................... 140

x

1
1.1

Work, Family, and the Organization
Introduction
Individuals in modern society assume many roles (V. Allen & van de Vliert, 1984; Katz

& Kahn, 1978; Stryker, 1980). Arguably most prominent of these are the work role and the
family role. Traditionally, these roles were largely split along gender lines with men focusing on
working and women focusing on taking care of family needs. But in the modern world, these
gender roles have substantially eroded (van der Lippe, 2010). Other family and work role
structures are becoming increasingly common as women contribute more to household incomes
(Raley, Mattingly, & Bianchi, 2006). For example, in 2008, 80% of married employees reported
being a member of a dual earner couple as opposed to 66% in 1977 (Galinsky, Aumann, & Bond,
2009). Female labor force participation rates have steadily climbed from 42% in 1950 to 57% in
2007 while male labor force participation rates have fallen from 82% to 66% in the same period
(Galinsky et al., 2009). Similarly, men now report increased performance of family role tasks;
for example, the ratio of women's to men's time spent on housework has dropped from 6.1 in
1965 to 1.6 in 2010 (Bianchi, Sayer, Milkie, & Robinson, 2012).
This still ongoing fundamental shift in work and family roles is not without
consequences. Time and energy are finite resources, but each role requires both. Thus, these
roles are often in competition with one another for effort and attention. Many people struggle to
be successful in both the work and family domains simultaneously. Work role demands can
interfere with the fulfillment of family role demands, and family role demands can interfere with
the fulfillment of work role demands, creating work-family and family-work role conflict,
respectively (Greenhaus & Beutell, 1985; Kossek, Pichler, Bodner, & Hammer, 2011). Similarly,
moods and behaviors from one domain can spill over into the other (Z. Chen, Powell, &
1

Greenhaus, 2009; Ilies et al., 2007). As a result, younger labor market participants are reporting
work-family role management as a key concern for their future careers (Burke, 2004; Cleveland
& McCarthy, 2013) and work-family matters are becoming more important in consideration of
career development (Greenhaus & Kossek, 2014). In this regard, workplaces that adopt official
written policies and less formal practices that help employees manage their work and family
roles are more desirable to today’s job applicants (Carless & Wintle, 2007; Hill et al., 2008; Rau
& Hyland, 2002; Rothbard, Phillips, & Dumas, 2005; SHRM, 2010).
Organizations are not blind to these broad societal changes, and many have chosen to
address the work-family needs of their employees through their system of human resource
management policies and practices. In the United States, organizations are the key source of
work-family benefits for employees due to a lack of public policy surrounding work-family role
management relative to other countries (Galinsky, Sakai, Eby, Bond, & Wigton, 2010; Piszczek
& Berg, forthcoming). Still, as will be discussed in Chapter 3, public policy in other countries
does not always provide all the work-family benefits employees need or want. The organization
thus plays a major role in employees’ ability to manage work and family roles and its actions can
have a substantial and often beneficial impact on employees’ lives (Butts, Casper, & Yang, 2013;
Keeney, 2012; Kelly et al., 2008; Kossek, Hammer, Kelly, & Moen, 2014; Kossek, Pichler, et
al., 2011; M. D. Lee, Kossek, Hall, & Litrico, 2011).
Research on the work-family interface has gained substantial popularity over the past
several decades, coinciding with shifting gender roles. Work by Kanter (1977), Hall and Richter
(1988) and Orthner and Pittman (1986), discussed below, helped promote dialogue of what
organizations could contribute to employees’ management of work-family roles. As part of this
research, in the early and mid 1990s researchers developed several organization-level

2

frameworks from different disciplinary perspectives meant to explain how organizations respond
to employees’ need to manage multiple roles. These frameworks sought to explain, at a general
level, whether and why organizations adopted work-family policies and practices, despite
relatively little being known about such policies and practices at the time (Goodstein, 1994;
Kirchmeyer, 1995). Because these frameworks were created at the onset of the explosion of
work-family research, and subsequently do not speak with much detail to the various types of
work-family practices and their effects on employee role management, they have largely fallen
out of favor with modern research. Instead, focus has largely been on the individual employee
level. Individual level work-family research explains how employees use and experience specific
work-family practices and manage their work-family role boundaries, as well as how the work
and family roles influence one another (e.g., T. D. Allen, Cho, & Meier, 2014; Casper, De Hauw,
& Wayne, 2013; Chakrabarti, 2011; Kossek & Lautsch, 2012; Kossek, Ruderman, Braddy, &
Hannum, 2012; Kreiner, Hollensbe, & Sheep, 2009; Ollier-Malaterre, Rothbard, & Berg, 2013;
Valcour, Ollier-Malaterre, Matz-Costa, Pitt-Catsouphes, & Brown, 2011).
Even though we now know much more about how work-family policies and practices
affect employees and organizations, organization-level frameworks still have something to offer
modern research. Through focusing on the individual employee and specific work-family
policies, the literature has lost much of the context that explains why a practice or policy is
adopted, how it is implemented, how it affects employees, and whether it should be deemed
successful or unsuccessful. In other words, these frameworks help answer the question of why,
with such a multitude of options, an organization ends up with its own custom array of workfamily practices and policies. They help answer the question of why the same policy can have
vastly different effects in different organizations, or why it might be more valuable to one

3

organization than another. They help address the question of why policies and practices might be
unevenly applied within an organization and what might come as a result for that organization
and its employees. Though individual-focused research often acknowledges the importance of
organizational context by including variables such as work-family culture or perceived
organization work-family support (Kossek, Lewis, & Hammer, 2010; Odle-Dusseau, Britt, &
Greene-Shortridge, 2012; Powell, Francesco, & Ling, 2009; Wayne, Casper, Matthews, & Allen,
2013), work-family research lacks an organization-level framework that can explain the variation
in adoption and implementation of work-family policies and practices observed in real
organizations.
An updated organization-level framework has a number of potential benefits. I develop
such a framework and discuss these benefits in detail in Chapter 4. This framework helps
contextualize existing and future research in a broader institutional environment, explains how
work-family policies and practices fit better or worse with different organizational needs,
clarifies theoretical links from organization to individual, environment, and practice, and
accounts for variation in the way policies and practices are implemented within and between
organizations. More generally, a new organization-level framework is needed to account for the
complex patterns of work-family policies and practices offered in and across organizations. Such
a framework also raises a number of potential areas for future research about cross-level
relationships between organizations and individuals, how work-family policies are chosen and
implemented, and what organizational contingencies are important for work-family policy
effectiveness.
Thus, through this dissertation, I argue for a return to a higher-level, organizational
perspective on work-family policies and practices to supplement the wealth of more micro- and

4

meso-level literatures. In this first chapter, I will discuss and critique three organization-level
work-family response frameworks from the mid 1990s (institutional pressure, rational choice,
and agency). I address why they have received such little attention, and what we can yet take
from them to build additional theoretical foundations for future work-family research. I also
discuss why organization-level theorization is important in work-family research and how it can
help explain complex patterns of organizational work-family policies and practices within and
across organizations.
Chapters 2 and 3 consist of empirical studies typical of the modern micro and macro
work-family literatures, respectively. Aside from their more narrow contributions, these two
studies demonstrate the need for a more comprehensive framework of how and why
organizations develop and implement a particular system of work-family practices and policies.
In response to this need, Chapter 4 introduces an integrated organizational work-family
framework. In discussing this new framework I highlight its theoretical, empirical, and practical
benefits and how it might be used in future research. I also discuss how this perspective relates to
current research and how this can be leveraged into stronger, better contextualized research. I
begin with an in-depth description of the three organizational work-family response frameworks
and a brief history of the literature that led to their development.
1.2

Three Organizational Work-Family Response Frameworks

1.2.1 Development
The popularity of work-family studies followed closely the beginning of the shift away
from traditional gender roles in modern society. Kanter (1977) argued that organizations take
different forms in the way they treat employees’ work and family roles. She deconstructed the
“myth of separate worlds,” that work and family domains were inherently disparate and thus one
5

need not interfere with the other. Kanter argued that organizations typically took two response
forms, based on whether or not they acknowledged the link between work and family domains.
This relatively early work was designed to make an ethical, humanitarian appeal to organizations
consistent with increasing prevalence of women in the workplace and dual earner couples in the
labor market. It did not demonstrate the potential economic benefit of work-family supportive
policies and practices to organizations (Orthner & Pittman, 1986; Voydanoff, 1984). Because of
this humanitarian grounding, most organizational work-family research during this period was
focused on the impact of work on family-domain outcomes such as marital quality and family
stress (e.g., Berk & Beck, 1979; R. Clark, Nye, & Gecas, 1978; Olson et al., 1983; Orthner &
Axelson, 1980).
In the early 1980s, research began to test whether organizational practices and policies
that acknowledge and support an employee’s non-work life could also benefit organizations. For
example, Orthner (1980) demonstrated that spousal support was related to career commitment in
men in the military. Following this, Orthner and Pittman (1986) called for more empirical
research linking work-family supportive practices and outcomes desirable to employers while
empirically demonstrating that organizational support for the family was linked to higher levels
of organizational commitment. Once empirical groundwork had convincingly demonstrated the
existence and importance of cross-domain relationships, researchers sought to understand what
organizations were doing—and what they could be doing—to allow the work and family
domains to maximally benefit one another. Hall and Richter (1988), building from Kanter’s
work, argued that while some organizations are unwilling to address work-family issues, others
had the desire but lacked knowledge on what types of practices would be mutually beneficial for
them and their employees. These authors identified a number of employee work-family needs

6

and ways in which organizations might be able to address them, prompting more research on
organizational work-family policies and practices. This led to the development of three
theoretical frameworks that sought to explain when and how organizations respond to the fact
that employees have multiple life roles.
1.2.2 The Institutional Pressure Framework
One approach to research on the adoption of organizational work-family policies and
practices is to treat the need to adopt work-family policies and practices as an institutional
pressure. Literature using this approach applies institutional theory to the work-family interface
at the organization level. Oliver (1991) discussed in depth the role of institutional pressure in
contributing to an organization-level response to that pressure. This research serves as a baseline
for studies of institutional pressure in the work-family domain. For example, Milliken (1990)
argued that the key drivers of an organization’s response to institutional pressures are the
organization’s external visibility and the salience of the potential consequences of work-family
problems for employee performance. However, Goodstein (1994) integrated Oliver’s (1991)
theoretical framework of strategic choice in responsiveness to institutional pressures into the
work-family domain, describing what high and low work-family institutional pressure might
look like.
Goodstein thus provides an in-depth discussion of what comprises work-family
institutional pressure using five predictive factors described by Oliver (1991): cause,
constituents, content, control, and context. “Cause” pressure includes the belief that
organizations should play a more socially active role in society by helping their employees; the
root of this pressure is in the organization’s need for social legitimacy, which would be impeded
if it did not conform to this pressure if present. Goodstein (1994) argued that larger organizations
7

would experience this pressure more than smaller organizations. “Constituent” pressure stems
from the demographics of the labor force. A greater proportion of the labor market in need of or
demanding work-family policies and practices (for example, a greater proportion of women of
childbearing age) places greater pressure on organizations to adopt them. “Content” pressure
comes from the consistency and congruence of work-family policies with other organizational
goals and policies. For example, Goodstein (1994) argued that public sector organizations would
be more likely to adopt work-family practices because they are more consistent with public
policy promoting work-family policies and the philosophy of government service. “Control”
pressure comes from legal regulations mandating particular work-family policies (such as the
1993 Family Medical Leave Act in the United States) and voluntary diffusion (i.e., pressure to
adopt practices that are widely adopted by other organizations). Finally, “context” pressure is
highest when organizations are highly interconnected. Goodstein (1994) argued that professional
and other business networking organizations in an institutional environment create pathways for
normative rules to promulgate across organizations.
Goodstein’s (1994) work also shows that organizations may take one of five responses to
the institutional pressure to adopt work-family systems described above depending on the
strength of the pressure and the perceived effect their adoption may have on technical outcomes
like company performance. These responses are presented in Table 1.1. Organizations may
acquiesce to the pressure and adopt work-family practices when the pressure to adopt is high and
the company may expect benefits from their adoption. Acquiescence is full conformity to
institutional pressures. When pressure is high but expected outcomes are negative, organizations
may compromise in their response or avoid the pressure completely if such avoidance carries
little risk. Compromise is partially complying with institutional demands while avoidance is

8

characterized by concealed non-conformity or token symbolic policies. When pressure to adopt
is low but there may be benefits to adoption, organizations may use manipulation to aggressively
adopt policy and exert power over the weak institutional context and gain legitimacy.
Manipulation is an active attempt to change or dominate institutional pressures to the
organization’s advantage. Finally, when pressure is low and undesirable outcomes from adoption
are expected, organizations may actively reject the notion of adoption in a defiance response.
Defiance is an active rejection of institutional norms or expectations.
Table 1.1. Goodstein’s (1994) framework for strategic responses to institutional pressures
Positive perceived effects of
Negative perceived effects of
responsiveness on technical
responsiveness on technical
outcomes
outcomes
High strength of institutional
Acquiescence
Compromise/Avoidance
pressure
Low strength of institutional
Manipulation
Defiance
pressure
Later empirical research provides some support for this framework, showing that industry
norms and expected impact on performance by executives were important contributors to the
adoption of work-family policies and practices (Milliken, Martins, & Morgan, 1998). Similarly,
organizational size and workforce demographics were found to be consistent predictors of
strategic response consistent with hypotheses grounded in an institutional pressure perspective
(Bardoel, Tharenou, & Moss, 1999; Ingram & Simons, 1995). This framework was also applied
to organizational eldercare provisions (Goodstein, 1995). Overall, though, the institutional
pressure perspective has not received much empirical attention.
1.2.3 The Rational Choice Perspective
Human resource management and economics literatures also generated explanations as to
why organizations would adopt work-family policies and practices. In these literatures, the
9

adoption of work-family practices can be considered a rational choice for efficiency and will be
adopted when their benefits exceed their costs (Glass & Estes, 1997; Poelmans & Sahibzada,
2004; Seyler, Monroe, & Garand, 1995). The notion of efficiency is somewhat captured in the
institutional pressure perspective, as Goodstein’s (1994) model includes the organization’s
expected effect of work-family policy on technical outcomes and Oliver (1991) discusses
efficiency as a component of the decision to respond to institutional pressure. However, this
literature takes a more nuanced perspective. While the institutional pressure literature simply
asks whether an organization expects work-family practices to help or hurt their bottom line, the
rational choice perspective seeks to better understand when and why an organization might
expect these outcomes. This perspective is most consistent with current strategic human resource
management literature (e.g., Huselid, 1995).
Osterman (1995) described three potential reasons for adopting work-family practices
based on human resource management concerns. Organizations might adopt work-family
policies in response to workforce problems like absenteeism, turnover, or labor market pressure;
as part of a well-developed internal labor market; or as part of a high-commitment work system
designed to increase employee organizational commitment. This literature also identifies unions
and organizational size as drivers of work-family policy adoption not as institutional pressures
but rather as strategic decision points for organizations (Glass & Fujimoto, 1995; Seyler et al.,
1995). The ideas of the rational choice perspective, despite its lack of a formal framework like
that of the institutional pressure perspective, are still present in current research though not
generally acknowledged as such (e.g., B. Lee & DeVoe, 2012).

10

1.2.4 The Agency Perspective
Building on the idea that organizations have an active role in determining their workfamily policies and practices, Kirchmeyer (1995) next formally defined the employer’s general
work-family responses in line with those identified by Kanter (1977). The previous models
posited by institutional pressure and rational choice perspectives were focused on the workfamily response in the form of the mere presence of work-family practices without putting much
emphasis on their content, their effects on employees, or the organizational philosophies they
represent. Kirchmeyer’s conceptualization of the organizational work-family response thus
provided a richer description of how organizations might differently address the work-family
needs of employees, though it did not provide much information about the source of these
differences.
The response types developed by Kirchmeyer are presented in Table 1.2. The “response”
column contains Kirchmeyer’s three responses, as defined by her in the following column. The
third column contains the organizational philosophy associated with each response in terms of
how the organization views its role in helping employees manage work-family demands. The
final column includes examples provided by Kirchmeyer of how organizations might enact each
response through policy. Because this perspective personifies the organization as having a
philosophical approach to the work-family interface (i.e., as paternalistic, mutualistic, or absent),
I refer to it as an agency perspective.
According to the agency perspective, a "separation" response entails a lack of any kind of
action to help employees manage work and family roles. Under this response, the organization
concerns itself only with employees’ fulfillment of the work role and leaves employees to
manage their family roles at their own discretion. Employees may be expected to keep their non11

work roles out of the workplace. It is worth noting that this type of response would still be
required to meet minimums of federal and state regulations, such as compliance with the Fair
Labor and Standards Act or Family-Medical Leave Act in the United States.
Table 1.2: Kirchmeyer’s (1995) work-family response archetypes
Response
Definition
Characteristic

Separation

Integration

Respect

No consideration of
employees’ workfamily needs.

Absent

Takes care of
employees’ nonwork needs

Paternalistic

Gives employees
tools to manage nonwork needs on their
own



Mutualistic

Examples

Organization prefers to
keep work and nonwork as
separate
Concerned mainly with
employees fulfilling work
role




Provides onsite daycare
Offers family/marital
counseling



Flexible about work
schedules
Avoids intruding on
personal time
Supports alternative work
arrangements




With an "integration" response, the organization recognizes the impact work and family
roles can have on one other and contributes in some sense to employees' management of nonwork roles. This occurs most typically through work-family policies, practices, or other
mechanisms which employees can use to help manage conflicting role demands such as onsite
childcare centers or company-provided family counseling facilities. Kirchmeyer describes the
motivation for the integration response as fitting a paternalistic model in which the employee
benefits from the more powerful employer’s protection; in this sense, the organization is helping
the employee to address his or her non-work needs more directly.
Finally, Kirchmeyer adds a third work-family response archetype based on the work of
Hall and Richter (1988): “respect.” Under this response, employers acknowledge, value, and
12

support employees’ non-work roles and provide employees with the tools they need to handle the
demands of these roles themselves, primarily through flexibility. This is in contrast to
integration, in which the employer assumes direct responsibility for meeting employees’ nonwork role needs. While the integration response is considered paternalistic, the respect response
is considered mutualistic; the ultimate responsibility for managing roles falls on the employee.
Kirchmeyer (1995) found that this response type yielded the best outcomes for employees,
including organizational commitment and positive spillover from non-work to work role
domains. Following research found that respect was associated with higher organizational
commitment, separation was associated with lower organizational commitment, and integration
was unrelated to organizational commitment (A. Cohen, 1997). Though some concepts of this
framework are similar to those of boundary theory (Nippert-Eng, 1991), a popular theoretical
grounding of many individual-level work-family studies, this framework has not received much
empirical attention.
1.3

Strengths and Weaknesses of the Organizational Work-Family Response
Frameworks
Each of the three frameworks remained largely unutilized in the modern work-family

literatures, though many of their core ideas are subtly present in studies with organizational
components (e.g., in strategic human resource management, work-family climate, and multilevel empirical studies). The lack of success of these frameworks is therefore not necessarily due
to a lack of useful ideas. At least part of their lack of success is attributable to the relative youth
of the field. At the time these frameworks were introduced, there were neither well-accepted
typologies of work-family practices and policies nor a detailed understanding of how they affect
employees and organizations. This was acknowledged by the developers of these frameworks at
the time. In testing her response type framework, Kirchmeyer (1995) clustered several
13

organizational practices into her response types, forming a typology of general organizational
responses to employee work-family needs. She argues that this typology helps to “conceptualize
and examine the array of practices” employers might adopt. But she also acknowledged that the
typology was limited by the relatively young body of research. Similarly, Goodstein (1994)
called for more research examining organizational and environmental factors which affect
organizations’ decisions whether or not to adopt policies and practices.
Another important shortcoming of these frameworks—and more likely the reason they
have fallen into disuse—is that they treat the organization's work-family response too
holistically. Organizations are dynamic, and it is difficult to argue that the organization has a
general attitude or overall response to employees' work-family needs. In other words, these
frameworks do not allow for much differentiation in the implementation of work-family policies
and practices within and across organizations. On the contrary, by classifying what might be
large and very diverse organizations with complex and uneven work-family practices into broad
categories, these frameworks assume that there is little room for variation in how work-family
policies and practices are implemented and applied. Yet, these frameworks still bring valuable
insights about organization-level phenomenon that affect the way organizations ultimately
implement work-family policies and practices. Though they do not necessarily line up perfectly
with existing theory and typologies of work-family practices, there is still much to be gained
from them.
The institutional pressure perspective provides context around the general decision of
whether or not to adopt a work-family policy or practice. This context is useful for understanding
the rationale behind a policy's adoption, the organizational goals with which a policy should be
most closely aligned, and explaining some variation in the prevalence of various work-family

14

policies and practices. However, this perspective does not provide much detail about the policies
themselves. For example, in empirically measuring an organization's practices, Goodstein (1994)
used a count model which treated childcare and flextime practices as equal and gauged an
organization's response on whether they offered one, both, or neither of these practices. But
practices should not be treated as equal as they may have different effects and structures.
Research has identified many types of work-family practices and classified them under various
different typologies, for example as providing flexibility in the timing, place, or amount of work
(Kossek & Michel, 2011) or as helping employees segment or integrate their work-family roles
(Rau & Hyland, 2002; Rothbard et al., 2005).
The rational choice perspective emphasizes the employee and organization-level effects
that work-family practices might have. While many studies over the years have acknowledged
the potential strategic value of work-family policies and practices for employee attraction,
retention, and sometimes even productivity (Batt & Valcour, 2003; Carless & Wintle, 2007;
Honeycutt & Rosen, 1997; Kossek & Nichol, 1992; B. Lee & DeVoe, 2012; Rau & Hyland,
2002; Richman, Civian, Shannon, Hill, & Brennan, 2008), much remains to be discovered about
when, how, and why work-family policies and practices benefit an organization's bottom line and
when this benefit is substantial enough to warrant their cost. For example, it is unknown which
types of practices appeal to which labor force segments and how heavily work-family policies
weigh in to application decisions. This perspective can contribute to the need to make the
business case for work-family (Kossek, Baltes, & Matthews, 2011; Kossek & Friede, 2005) and
to an understanding of why work-family initiatives frequently fail or are marginalized by upper
management (Kossek et al., 2010). This contribution is important because critics accuse the
strategic human resource management perspective of ignoring institutional context and assuming

15

positive effects on organizational performance despite the context (Kaufman, 2010).
Additionally, much like the institutional pressure perspective, the relatively small body of
rational choice literature does not include much detail about the types of work-family policies
and practices or when they might be most cost-effective.
The agency perspective's advantage is that it includes more information about how
policies and practices might work and the organizational philosophy that might drive them. The
three responses in this perspective (respect, integration, and separation) include a component
describing how organizations manage employee work-family needs, whereas the others focus
more generally on whether there will be a response at all. But this framework has problems as
well.
While holistically categorizing complex organizations into only three types is
problematic as discussed above, the three categories offered by this framework are also not
consistent with current work-family research. For example, an organizational separation response
does not distinguish between organizations that ignore the work-family needs of employees and
those that encourage employees to actively segment their work and family roles, keeping them
from interfering with one another as much as possible. Current research suggests that segmenting
is a viable boundary management strategy for individuals and that policies and practices that
grant employees greater boundary control to manage their roles as they wish are associated with
better employee outcomes (e.g., Kossek & Lautsch, 2012; Kreiner et al., 2009). The
conceptualization of the separation response does not recognize the difference between policies
and practices that give employees greater boundary control to segment roles and the absence of
any such practices. Additionally, these categorizations do not align neatly with work-family
policies and practices. Researchers attempting to use this framework have been unable to agree

16

on whether a particular practice such as flextime or childcare should be considered as
encouraging segmenting, integrating, neither, or both (e.g., Poelmans & Beham, 2008; Rau &
Hyland, 2002; Rothbard & Ramarajan, 2009). These categorizations require major updates to be
consistent with current work-family research and useful in understanding how work-family
policies and practices are chosen and implemented by organizations.
A fully integrated model of the organizational response to employee work-family needs
was not possible when this organization-focused line of research was developing. Indeed, these
three different ways of looking at the work-family organizational response (institutional pressure,
rational choice, and agency) all emerged nearly simultaneously and without much consideration
of one another or actual practices and policies that have aptly been the focus more recently.
Given the substantial growth of work-family research over the past 20 years in this area, many
improvements to these frameworks are possible that can retain their strengths while attenuating
many of their weaknesses.
While individual-level research continues to remain important and evolve as labor force
demographics, regulative policy, and even the fundamental way work is performed all continue
to change, much of our current knowledge remains unsynthesized and over-generalized. For
example, researchers generally assume that work-family practices should help organizational
attraction and retention, but organizations may offer work-family policies for many different
reasons including as part of a human resource strategy, a corporate social responsibility
movement, or merely to comply with the law (Huselid, 1995; Jones & Murrell, 2001; O’Connor,
2005; Piszczek & Berg, forthcoming). The reason for the adoption of policy may contribute to its
implementation, which is understood to have a substantial impact on a policy’s effectiveness
(Ryan & Kossek, 2008). Similarly, employee attributions of a policy or practice to an entity may

17

affect how it shapes their attitudes toward that entity (Nishii, Lepak, & Schneider, 2008). The
same policy on paper can thus mean many different things across organizations. In order to
understand which policy best fits the needs of an organization and the context in which policies
are effective or ineffective, more organizational and multi-level theoretical foundations are
needed.
As I will discuss in greater detail in Chapter 4, a more complete framework that
integrates the strengths of these three frameworks while remaining consistent with existing
individual-level research has numerous advantages for future research. Such a framework would
first help researchers better understand why organizations offer—or do not offer—particular
work family practices and policies. Second, it would help explain which practices and policies
contribute best to which broader organizational goals. Third, a more complete framework would
be easier to align with modern work-family research that emphasizes practices and their effects
on employees, and could explain consistency in the adoption of practices both within and across
organizations. Fourth, though multi-level models can help put boundaries on the results of more
micro-focused studies, a new organizational response framework can help researchers better
contextualize their samples and identify which organizational and environmental factors might
be most important to consider as study boundaries or as key variables. Fifth, such a framework
provides a theoretical grounding for multi-level work-family studies and allows for complex
patterns of work-family policies and practices. While the existing organizational work-family
response frameworks are largely outdated or unadopted, many of their ideas (e.g., motivation, the
legal context, strategic value) are still important and draw attention to the need for a broader
perspective in work-family research. An integrated framework as described above has the

18

potential to explain inexplicable complexities in actual organizational practices and fuel new
multi-level research.
1.4

Dissertation Structure
The remainder of this dissertation is written with an organizational perspective on the

work-family interface in mind. Having introduced and critiqued existing organization-level
frameworks meant to explain generally how and why organizations offer work-family practices
and policies, two goals remain. The first goal is to present both micro and macro-level original
empirical work-family research examining under-researched aspects of the organizational
context that contribute to current research while providing evidence for the need for an integrated
work-family organizational framework. The second goal, using these studies and other current
research, is to develop such a framework and demonstrate its potential contribution to workfamily theory, empiricism, and practice.
In order to accomplish these goals, this dissertation is broken into four chapters. The
present chapter introduces the organizational work-family response literature, details the three
organizational work-family response frameworks, and provides a breakdown of their strengths
and weaknesses relative to modern research. Chapters 2 and 3 each consist of a standalone
empirical study related to organizational work-family practice and policy important to current
research topics but also the development of an integrated framework.
Chapter 2 focuses on the area of informational and communicative technology in
knowledge-based work, a topic of growing importance and popularity. Research suggests that
use of such technology is tied to role identities and can result in either increased flexibility in the
time and place of work or employees simply performing more work and risking exhaustion
(Derks & Bakker, 2012; Kossek, Lautsch, & Eaton, 2009; A. Ollier-Malaterre et al., 2013). I
19

argue that an organization’s expectations of employees’ use of technology will influence the way
that this work is experienced by employees, either as a job demand or as a job resource (Karasek,
1979). I leverage boundary theory (Ashforth, Kreiner, & Fugate, 2000; Nippert-Eng, 1991,
1996; Zerubavel, 1993) to show how the use of information and communicative technology can
be experienced as a job resource, and the technical control perspective (Callaghan & Thompson,
2001; R. Edwards, 1979) to explain how it can be experienced as a job demand. Furthermore, I
investigate how the actual use of information and communication technology and the pressure to
use it are experienced differently by employees with different boundary management styles,
building on both studies of how technology impacts work (Chesley, 2005, 2006; Derks &
Bakker, 2012; Diaz, Chiaburu, Zimmerman, & Boswell, 2012; Fender, 2010; Fenner & Renn,
2004, 2010; A. Ollier-Malaterre et al., 2013; Turel, Serenko, & Bontis, 2011) and how
individuals manage their work-family role boundaries (Ammons, 2013; Kossek, Lautsch, &
Eaton, 2006; Kossek et al., 2012; Kossek & Lautsch, 2012; Winkel & Clayton, 2010).
Using a sample of approximately 200 human resource managers whose work is amenable
to performance via information and communication technologies, I show that organizations’
expectations of employees to stay engaged with the work domain outside of normal work hours
contributes to employee stress when it is inconsistent with the employee’s preferred style of role
boundary management. The study uses data from a two-wave, self-collected survey consisting of
several scales to measure new constructs related to work-family role management and
technology. Using multiple regression and rigorous tests of mediation and moderation, I find that
employees that perceive higher organizational after-hours electronic communication expectations
generally experience higher emotional exhaustion and turnover intentions, but that this
relationship is especially strong for those who prefer to keep work and family roles segmented.

20

This study shows that technology should be considered as a tool for managing work-family roles
and that it is inherently neither helpful nor harmful to employee well-being. Instead, it is the
organizational pressure to use technology in certain ways that leads to greater employee stress.
The second study (Chapter 3) tests whether adoption of childcare practices by German
establishments has led to reductions in employee turnover. I discuss in depth the environmental
factors that make Germany a particularly appealing place for organizations to adopt childcare
practices to increase employee retention, highlighting the cultural interaction of work and
motherhood and public policy surrounding childcare. Recent research has argued that a lack of
public policy granting employees the ability to manage their work-family roles provides an
opening for organizations to step in and fill those gaps (Piszczek & Berg, forthcoming). I draw
upon strategic human resource management theory (Huselid, 1995) to argue that doing so allows
establishments the opportunity to reduce turnover levels.
Chapter 3’s analysis uses data from three years of the nationally representative,
government-collected IAB Establishment Panel, which includes surveys of approximately
16,000 establishments per year in a stratified random sample of German establishments. I show
that organizations that adopt childcare practices tend to have higher levels of turnover, but
experience a reduction in turnover as a result of adopting these practices. The pattern of results
suggests that, counter to initial expectations, establishments with high levels of turnover are most
likely to have childcare practices. But, using establishment fixed effects to examine only change
within establishments over time, the results show that establishments that adopt childcare
practices subsequently experience reduced levels of turnover.
This study demonstrates not only that work-family practices (particularly childcare) may
be able to provide strategic value to organizations but also that the contextual environment can

21

be very important in determining how to maximize this strategic value. Furthermore, this study
makes empirical contributions to the strategic human resource management literature by
separating within- and between-establishment relationships into separate analytic components,
helping to address controversy encircling the causal direction of the relationship between human
resource practices and organizational performance.
Finally, Chapter 4 discusses the results of the two empirical studies in the broader context
of the organizational work-family response frameworks introduced in the present chapter and
uses them to construct a new integrated organizational work-family response framework. I link
key concepts from each existing framework together and to constructs known to be important
based on current research. I show how each framework contributes different pieces to the puzzle
of how organizations respond to employee work-family needs. In building this framework I
make a case for the utility of an organizational work-family response framework in the modern
work-family research and explain how it can be used to further research and practice.
1.5

Conclusion
Overall, this dissertation seeks to shed some light on the many ways in which

organizations respond to employees’ need to manage work and family roles in a fashion that
allows for the complex patterns of practices and policies observable in real world organizations.
The two empirical studies take substantially different approaches, paralleling the vast spectrum
of ways that organizations are considered in the work-family literature. The first is focused on
the employee experience of an informal aspect of organizational work-family culture in a small
sample survey of HR managers in the United States, while the second is focused on
organizational performance of establishments in Germany using archival establishment-level
surveys collected by the German government. But despite their differences in scope and focus,
22

both speak to the same question: How do organizations react to a world where employees
increasingly need to devote time to non-work roles? Individually, these two studies provide some
specific insights into this question and contribute to more narrow work-family literatures: the
growing literature on technology and work-family role boundary management, and the strategic
human resource management literature grounded work-family policy literature. But together, as I
will discuss in Chapter 4, they show that the navigation of work and family roles is complex and
organizations can react to employees' work-family needs in many different ways, for many
different reasons, and with potentially many different outcomes.

23

2
2.1

Two Faces of Technology: Organizational Expectations at the Work-Family Interface
Introduction
The previous chapter discussed the development of the organizational work-family

response framework and its role in past and current research. Chapter 2 examines a specific part
of an organization’s work-family response that has evolved as organizations adapt to new ways
of working created by technological advances. The focus of this study is on the expectations that
organizations set for their employees’ use of information and communicative technologies,
which is one largely informal way an organization may attempt to manage employees’ workfamily needs. It is a stand-alone empirical study that provides major insight into a current issue
in the human resource management and work-family literatures, but also has implications for the
work-family response framework.
As I will explain in greater depth, researchers have identified conflicting effects of
information and communicative technology use on employee well-being. On one hand, it can
give them greater boundary control and help them work more efficiently, but on the other hand it
can lead to intensification of work and spillover of work into the family domain (e.g., Diaz et al.,
2012). This study answers the call for more research on how information and communication
technology affects employee boundary management (Day, Scott, & Kelloway, 2010) by further
developing and testing the recently created construct of after-hours electronic communication
expectations (Fender, 2010) and examining how these expectations shape the way different
employees use and experience information and communicative technology.
2.2

Technology at the Work-Family Interface
Rapid advances in technology are creating new issues in the world of work. In mid-

October, 2011, BlackBerry smartphone users across the world experienced a crisis when they
24

suddenly found themselves the victims of an international data network collapse. Because
professionals saw their BlackBerry smartphones as essential to their work, the backlash was
substantial. Some customers initiated a class-action lawsuit against BlackBerry’s network
maintainers and many others threatened to move to other service providers (Taylor, 2011).
Around the same time, Volkswagen reached an agreement with its German employees to reduce
the reach of technology by eliminating after-hours emails in response to complaints of blurred
work-family boundaries (BBC, 2011). Volkswagen barred work-related e-mails on Saturdays.
Thirty five percent of all adults in the United States owned a smartphone in 2011, rising
to 56% in 2013 (Smith, 2011, 2013). A recent poll suggests that up to 80% of working adults in
the U.S. take work home with them in some form, spending an average of 30 hours a month on
work tasks outside of the workplace (Good Technology, 2012). However, the same report also
suggests employees are working more on their own schedule. Information and communicative
technologies (ICTs) have changed the way that people work over the past few decades and
become a fixture in the modern workplace (Boswell & Olson-Buchanan, 2007). These
technologies consist of personal devices and services that facilitate the transmission of
knowledge by enabling written or vocal communication across physical and temporal
boundaries. ICTs such as e-mail, voice mail, text messages, and instant messages are sometimes
discussed interchangeably with their associated delivery mechanisms, such as laptop computers
or smartphones.
As I discuss in depth below, ICTs are thought to be used by organizations to improve the
flow of information, streamline organizational processes, and simplify coordination processes.
But while ICTs may provide such benefits, they may have other less desirable effects on
employees. As these technologies continue to advance, so too does the ability to communicate

25

with the workplace outside of the normal work domain and perform work tasks in practically any
physical location and time (Kreiner et al., 2009). Identifying, understanding, and appropriately
balancing the potentially conflicting effects of ICTs remains a challenge for work-family
researchers and human resource managers.
Understanding the implications of employees’ adoption and use of ICTs is most critical
for employees in knowledge management roles. Organizations have been increasingly using
ICTs for the purposes of knowledge management (Schultze & Boland Jr., 2000) and knowledge
work is simultaneously becoming more dependent on technological systems (Bailey, Leonardi,
& Chong, 2010). This is critical as Organization for Economic Cooperation and Development
(OECD) country economies grow more dependent on the transmission of knowledge (OECD,
1996). Knowledge-based workers’ core job tasks involve the creation and transfer of knowledge
and they are often at the forefront of new ICTs. While other types of employees may use ICTs
frequently for purposes of coordination or communication (e.g. to try to trade shifts with
someone at a manufacturing facility or to ask a colleague whether a certain task has been
performed), knowledge-based workers add value to an organization primarily through creation
and transmission of knowledge (Blackler, 1995). Because knowledge-based work has become
more important to organizational performance and has grown in tandem with technological
advances, knowledge-based workers are an ideal starting point to examine the impact of ICTs on
employee well-being.
Another reason to focus on knowledge workers is that although knowledge work is
arguably most amenable to performance via ICTs, the implications of using ICTs for such work
are poorly understood. Both researchers and human resource management practitioners have
struggled to keep up with rapid technological advances and their adoption by the working world,

26

even though core job tasks of researchers and human resource management practitioners
themselves are often performed through ICTs.
Partially due to the lack of concrete empirical studies, the use of ICTs by employees has
been controversial both in practice and in research. While some praise the flexibility they grant
employees, others suggest that they have intensified and lengthened the workday, with
undesirable consequences for other life roles. Practitioners and researchers alike do not know the
full extent of these consequences of ICT use for employees or how they can help employees use
ICTs in ways which help them work smarter rather than longer. In order to maximize the utility
of ICTs, employers need to understand how to build ICTs into the design and organization of
work in a way which does not put employees at risk of overwork. As of now, employers’ main
concerns around ICT use seem to be elsewhere.
A recent survey by the Society for Human Resource Management (2012) suggests that
the majority of employers that have after-hours ICT use policies craft them around data security
and privacy rather than employee well-being. That is, if they have ICT use policies at all.
Arguably as a result of the lack of knowledge around the consequences of ICT adoption in the
modern working world, only 21% of organizations have a formal policy for after-hours ICT use.
More reported having an informal policy (26%). Given the growing adoption and use of such
technologies and potential legal ramifications (such as whether ICT-based work in the home
could be considered overtime) (Kakabadse, Porter, & Vance, 2009), it is critical for
organizations to understand the impact of ICTs on their employees and create policies to help
manage that impact into desirable outcomes for both employer and employee. Of the 21% of
organizations with formal policies in place, 53% were concerned only with company-owned
devices provided to employees while only 27% mentioned a concern with work-life balance.

27

This report also finds that of those organizations without formal or informal policies surrounding
wireless communication device usage, 87% allow employees to manage their use on their own.
Thus organizations’ most common response to issues of ICTs is to let employees manage
ICTs—and their effects on work-family role management—on their own. However, research
shows that proactively considering employees’ work-family needs is associated with increased
employee well-being and organizational commitment (Bulger, Matthews, & Hoffman, 2007;
Hall & Richter, 1988; Orthner & Pittman, 1986). Furthermore, organizations that perceive
themselves as having no work-family agenda may in fact be informally encouraging the use of
ICTs for work purposes at home. As this study will show, employers play a critical part in the
way employees manage ICT use through informal expectations, a component of organizational
culture. Understanding the part of organizations in employee work-family role management is
essential to crafting quality ICT use policies that take into consideration employees’ role
management and well-being but still meet organizational needs.
Such an understanding is lacking not only in practice but also in research. Despite being a
topic receiving more and more attention, scholars lack a clear understanding of when and how
ICTs are helpful or hurtful to employees. Collectively, studies have yielded mixed results.
Streams of research arguing the promises and pitfalls of ICT use have recently converged, as
current research has begun to explore why in some cases ICTs are helpful to employees and in
other cases they are hurtful. One explanation may be that employees are better able to appreciate
the flexibility benefits of ICT use when they are able to use ICTs freely, i.e., without being
pressured by their employers to use them in a certain way. As I argue below, organizations that
have organizational expectations keeping employees electronically tethered to the workplace

28

after normal working hours may be putting some of their employees at risk of increased
emotional exhaustion, potentially leading to turnover.
Furthermore, this body of research has not adequately captured the role of employee
characteristics in examining the relationship between ICT use and employee outcomes.
Individual characteristics such as an employee’s work-family role management style may play a
critical part in determining how employees experience expectations of ICT use and ICT use
itself. Certain individuals may be more at-risk for problems associated with ICT use than others.
Similarly, certain individuals may more readily experience the benefits of ICT use than others.
Yet what individual characteristics determine these differences remain unknown.
In the present study, using a two-wave survey of 204 human resource managers, I show
that ICT use is neither inherently good nor bad; it is the reason for its use that largely determines
whether employees use it to work smarter or work more. This study thus contributes to theory in
multiple ways. First, it contributes to the emerging body of literature bringing together
competing views on ICTs and their conflicting effects. Second, it identifies both organizational
and individual-level factors which may influence how employees use and experience technology.
Third, it adds considerable theory from multiple disciplines to what has largely been treated as an
empirical question by utilizing the job demands-resources model (Karasek, 1979), boundary
theory (Ashforth et al., 2000; Nippert-Eng, 1991, 1996), and the technical control perspective of
the organization (R. Edwards, 1979). Finally, it demonstrates the importance of considering an
organization’s expectations for employee ICT use in studying work-family role management in
the modern workplace.
Practically, the study also shows that even though organizations mostly let employees
manage their own ICT use on their own time, they may be having an unconsidered and

29

undesirable influence on how and how much employees use ICTs. I argue that formal policy and
clear expectations are better ways to ensure ICT use by employees that avoids creating
unnecessary strain and risk of turnover, while still allowing employees to experience the other
benefits of ICT use such as greater flexibility and enhanced coordination. The remainder of this
chapter reviews the literature surrounding the purported desirable and undesirable effects of ICT
use for employees, leverages the three above-mentioned theoretical perspectives to combine
these two empirical streams and identify the underlying reasoning for their conflicting findings,
and then describes and tests the resulting hypotheses.
2.3

The Two Faces of Technology
Most research on the use of information communicative technology for work purposes

while at home falls under one of two perspectives: treating it as an intensifier of work or as a
flexibility enabler. The intensification perspective outlines several mechanisms through which
ICT use can be difficult for employees managing multiple role demands and can lead to
increased role conflict and stress. One emerging perspective suggests that the use of ICTs for
work purposes at home leads to stress by preventing post-work recovery (Derks & Bakker,
2012). However, the two most well understood mechanisms through which ICTs are thought to
intensify work are first by increasing the overall total amount of time employees spend working
and second by allowing work demands to interfere with meeting of demands from other life
domains (Barley, Meyerson, & Grodal, 2011; Duxbury, Towers, Higgins, & Thomas, 2006;
Fligstein & Sharone, 2002; Towers, Duxbury, & Thomas, 2005). These two mechanisms can
each be explained by separate theoretical perspectives related to working time.
The first mechanism linking ICT use to work intensification states that it can increase the
overall amount of work an employee performs, fitting more work tasks into the same amount of
30

time or spending time and effort on work tasks outside the work domain. E-mails are faster,
easier, and cheaper than other forms of communication (Bellotti, Ducheneaut, Howard, Smith, &
Grinter, 2005; Dawley & Anthony, 2003; Manger, Wicklund, & Eikeland, 2003; Thomas et al.,
2006). They allow employees to communicate more effectively (Dedrick, Gurbaxani, &
Kraemer, 2003) and can be used to strategically improve employee productivity (Baron, 2005).
The work-intensification perspective is consistent with Edwards’ theory of technical control of
employee effort (R. Edwards, 1984), which states that organizations maximize employee effort
through the design of technical systems in the workplace. Though this perspective was formed
with manufacturing production flow in mind, it still applies today with more modern technical
systems and has been applied to research in modern workplaces such as call centers (e.g.,
Callaghan & Thompson, 2001). ICTs allow employers to have higher expectations for the
amount of work accomplished in one day. By manipulating the pacing and flow of work through
requiring increased use of ICTs, employers can exert more total work effort from employees.
Similarly, ICTs allow employers to have higher expectations regarding availability of
employees after work hours. This is the basis for the second mechanism linking communicative
technology to work intensification: ICTs change the boundaries between work and non-work
roles, enabling work-domain tasks to more easily spill over into other life domains (Boswell &
Olson-Buchanan, 2007; Chesley, 2005; Duxbury et al., 2006; Major, Klein, & Ehrhart, 2002;
Towers et al., 2005). This spillover and its importance are explained by boundary theory.
Boundary theory (Nippert-Eng, 1991, 1996; Zerubavel, 1993, 1996) posits that
individuals vary on a continuum in the degree they prefer to integrate or segment their various
life roles. They construct mental boundaries separating each role domain, which can vary in the
degree to which they are flexible (i.e., can be physically or temporally adjusted) and permeable

31

(i.e., the ease with which a role domain's mental space can be infiltrated by another's). Role
domains with flexible and permeable boundaries are easily integrated while those without them
are easily segmented. There are tradeoffs inherent with each type of strategy (Keeney, 2012).
Roles with flexible bounaries are easier to transition between but may blur together (Ashforth,
Kreiner, & Fugate, 2000). Research linking ICTs to work intensification shows that ICTs enable
work demands to invade the non-work domain (Jacobs & Green, 1998; Schor, 1993). Similarly,
ICTs can also create interruptions in concentration and create additional coordination and time
management tasks, which may increase task completion time (Bellotti, Ducheneaut, Howard, &
Smith, 2003; Bellotti et al., 2005; González & Mark, 2004; Renaud, Ramsay, & Hair, 2006;
Whittaker & Sidner, 1996).
Thus, ICTs have the potential to make employees work harder and longer, and allow
work domain tasks to more easily invade other life domains. But ICTs are also thought to be
helpful for employees. Though flexible and permeable role boundaries can lead to blurring and
interruptions, they also have beneficial aspects. ICTs create role boundary flexibility, which is
associated with easier transitions from work to non-work role domains and vice versa (Ashforth
et al., 2000). While the flexibility granted by communicative technology makes it easier for work
demands to invade the non-work domain, it also makes switching back and forth rapidly between
demands easier. By allowing employees to work more flexibly, ICT use can potentially grant
employees additional autonomy, control over their work location, and customization of working
time. For instance, ICTs have been argued to grant employees control over their work-life
boundaries to integrate or segment role demands as necessary (Golden & Geisler, 2007;
Kaufman-Scarborough, 2006; Robinson & Godbey, 1997). Indeed, some research suggests that
employees tend to see ICTs as an asset rather than a burden (Baron, 2005; Day, Scott, &

32

Kelloway, 2010). Research also shows that control over work-family boundaries is linked to
desirable employee outcomes (Kossek et al., 2012).
Additionally, many of the reasons that allow ICTs to intensify work also allow them to
make work easier. As mentioned above, e-mails are faster, easier, cheaper, more effective, and
more productive than other forms of communication (A. Baron, 2005; Bellotti et al., 2005;
Dawley & Anthony, 2003; Dedrick et al., 2003; Manger et al., 2003; Thomas et al., 2006). While
the relative ease of ICT-based work allows organizations to raise the bar for the amount of work
performed in a given time period, it can also mean work requires less time and effort than in the
past working world without them.
All in all, there is substantial empirical evidence and theoretical rationale suggesting that
ICT use can have both potentially desirable and undesirable effects on employees. Recently,
research has begun to move toward a more nuanced perspective that differentiates between these
competing effects (Day et al., 2010; Derks & Bakker, 2010; Diaz et al., 2012; Matusik & Mickel,
2011; Nansen, Arnold, Gibbs, & Davis, 2010). Some studies in this newer line of research on the
dual-centric nature of ICTs have drawn from a job demands-resources perspective (Day et al.,
2010; Derks & Bakker, 2010).
The job demands-resources model (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001;
Karasek, 1979) posits that stress and exhaustion on the job are a result of two processes: high job
demands or low job resources. Job demands are “physical, social, or organizational aspects of the
job that require sustained physical or mental effort and are therefore associated with certain
physiological and psychological costs” while job resources are those that either are “functional in
achieving work goals,” “reduce job demands,” or “stimulate personal growth and development”
(Demerouti et al., 2001). Resources are necessary for individuals to reach their goals; too many

33

demands can lead to exhaustion while insufficient resources can create withdrawal or
disengagement from the job and decreased motivation as a mental mechanism to prevent further
frustration.
This perspective helps contextualize the competing effects of ICTs by allowing
identification of situations in which they are either helpful or hurtful. In other words, ICTs can
function as either a demand or a resource (Day et al., 2010; Derks & Bakker, 2010). ICTs can be
a job resource when they help employees work more efficiently or increase productivity (i.e.,
being functional in the achievement of work goals) or when they help employees better manage
their work and family role demands (i.e., reducing the job demand of role conflict). ICTs can be
a job demand when they create role conflict or additional work pressure. The context of the use
of an ICT may help determine whether it primarily functions as a job resource or as a job
demand. A review of ICT research reveals that few such contexts have been identified and even
fewer have been empirically examined.
One relatively early study—with data from before the widespread adoption of
smartphones—discussed the conflicting effects of technology in a descriptive review of research
on ICT use focusing on email (Derks & Bakker, 2010). Day and colleagues (2010) largely
discuss the role of ICTs in the workplace rather than at home. They provide a rich discussion of
potential factors which may make communicative technology function more as a job demand or
resource, but do not test any of these factors. A qualitative study found that the reasoning behind
the use of ICTs played an important role in whether people reacted to it in positive, negative, or
mixed terms (Matusik & Mickel, 2011). A more recent study shows that perceptions that ICTs
provide flexibility were related to work satisfaction and work-life conflict, mediated by
technology use (Diaz et al., 2012). This study argues that those who perceive technology to help

34

them work more flexibly would use technology more and experience more desirable outcomes,
but it does not identify the conditions which enable employees to use technology flexibly or
could potentially link technology use to undesirable outcomes.
Only one study shows moderating conditions that determine whether ICT use has
desirable or undesirable effects on employees. Fenner and Renn (2010) identified several
antecedents to what they termed technology-assisted supplemental work, and showed that
individual time management skills helped reduce its positive relationship with work-family
conflict. This study also showed that organizational pressure to use technology for work
purposes outside of the workplace was related to its actual use. Overall, researchers acknowledge
that an understanding of the desirable and undesirable effects of ICT use and when and why each
occurs is important but lacking in existing research (Day et al., 2010; O’Driscoll, Brough,
Timms, & Sawang, 2010). So far few potential explanatory mechanisms to understand when
which effect—work intensification or work flexibility—might take precedence have been
identified and empirically tested.
As such, researchers have an incomplete understanding of when ICT use may function as
either a job demand or job resource. Organizational climate for ICT use is one contextual factor
which may play a major role in how ICTs are used by employees. In a theoretical paper, Fenner
and Renn (2004) leveraged theories of organizational behavioral norms (Rousseau, 1990) to
argue that organizational climate could create an increase in the use of technology-based work
outside of work hours. These authors later provided empirical support for this argument,
described above (Fenner & Renn, 2010). In their qualitative study, Matusik and Mickel (2011)
similarly identified this pressure as a key component in interviewees’ experience of technology
and a motivation of its use.

35

The current study builds upon this argument by suggesting that organizational climate
(specifically in the form of after-hours electronic communication expectations) plays a key role
not only in the extent of employees’ use of ICTs as shown in previous studies but also how
employees experience their use. I argue that after-hours electronic communication expectations
determine whether employees experience ICT use as a job demand or job resource. I also argue
that individual differences may also shape how these expectations are experienced by employees.
The current study draws heavily on boundary theory to build upon previous studies and explain
different patterns of technology use and why individuals in the same organization may not
experience expectations in the same way.
2.4

Hypothesis Development
The conceptual model for the present study is shown in Figure 2.1. As explained above, a

core tenet of boundary theory is that some people prefer to integrate their work-family roles
while others prefer to segment them. Though there is recent support for more complex and less
temporally stable patterns of role boundary management including additional categories such as
alternators (Kossek & Lautsch, 2012; Kossek et al., 2012), the majority of boundary
management research places individuals’ boundary management preferences somewhere on a
continuum between complete integration and complete segmentation (G. Chen, Kanfer, DeShon,
Mathieu, & Kozlowski, 2009; S. Desrochers, 2005; Stephan Desrochers & Sargent, 2004; J.
Edwards & Rothbard, 1999; Kreiner et al., 2009; Kreiner, 2006).
While ICTs can potentially be used to segment work and family roles (e.g. by turning off
a cell phone to avoid work-related phone calls or keeping separate work and personal calendars),
ICTs are primarily a tool of integration. They increase permeability of the family role boundary,
creating a pathway through which work role demands can spill over to the family domain. ICTs
36

increase flexibility of the work role boundary by making it possible to perform work in places
and times outside the typical sphere of the work domain. Because ICTs are primarily a tool of
integration, those who prefer integrating their work roles into the family domain should use
communicative technology more for work purposes while at home. Conversely, those who prefer
to segment their work role from the family domain should use ICTs less for work purposes while
at home. This is shown in Figure 2.1 with the arrow from work-to-family segmentation
preferences to work-to-family ICT use. In the model, I use “segmentation preferences” to
indicate the degree to which an individual prefers to segment work from family as opposed to
integrating work into family.
Figure 2.1: Theoretical model with hypotheses
WF
Segmentation
Preference
H1

WF Conflict
H5

H4

H2
AECE

H1

H3
Emotional
Exhaustion

WF Tech Use

Notes: Dashed lines represent negative relationships.
AECE=After-hours electronic communications expectations

H6

Turnover
Intentions
Emotional
Exhaustion

Hypothesis 1: Work-to-family segmentation preferences are negatively related to work-to-family
ICT use.
Still, even segmenters may sometimes feel compelled to integrate work role demands into
the family domain under certain circumstances. According to the technical control perspective
(Edwards, 1979), organizations can utilize technology to structure work in order to maximize

37

employee effort. Some organizations may have a social norm in which employees are pressured
to stay electronically tethered to the organization, being available to meet work role demands at
all times and places. Some research has shown that this pressure or social norm—called afterhours electronic communications expectations —is related to a perception of being electronically
tied to an organization (Fender, 2010), and similar climate measures have been linked to use of
ICTs for supplemental work purposes (Fenner & Renn, 2010). A segmenter in a high after-hours
communications expectation climate must still integrate work into the family domain in order to
conform to organizational norms and expectations and remain employed. These expectations can
override the segmenter's personal desire to segment, potentially erasing the difference in ICT use
behavior between segmenters and integrators hypothesized above. This moderating relationship
is shown in Figure 2.1 by the arrow from after-hours communications expectations pointing to
the arrow linking work-to-family segmentation preferences to work-to-family ICT use, indicated
by the label H2.
Hypothesis 2: The negative relationship between work-to-family segmentation preferences and
work-to-family ICT use is moderated by after-hours electronic communications expectations
such that the relationship is less negative when expectations are higher.
The Job Demands-Resources Model predicts that the absence of job resources should be
related to disengagement from work while the presence of job demands should be associated
with increased burnout (Demerouti et al., 2001; Schaufeli & Bakker, 2004). One dimension of
burnout, shown to be strongly linked to job demands, is emotional exhaustion (R. T. Lee &
Ashforth, 1996). Emotional exhaustion is a feeling of being overextended by the emotional
demands of work (Demerouti et al., 2001). It is widely treated as a measure of stress based on its
conceptual overlap with more direct stress reactions and relationship with stressors such as

38

workload, role problems, and stress outcomes like turnover intentions and absenteeism
(Demerouti et al., 2001).
Because ICTs in and of themselves are neither a job demand nor a job resource, their use
should not be directly linked to employee exhaustion. But, an organization's after-hours
electronic communications expectations may provide insight into whether ICTs are functioning
as a demand or a resource in a particular organizational context. A higher after-hours
communications expectation climate suggests that the use of ICTs for work purposes at home is
at least partially driven by organizational pressure rather than an ability to work more flexibly. In
this context, ICT use is experienced as a job demand that increasingly uses up an employee’s
resources or prevents resource recovery (Derks & Bakker, 2012). Employees using ICTs in the
work-to-family direction may end up feeling more exhausted or burned out with their work due
to the work intensification effect (i.e., working faster and working more overall).
On the other hand, when employees are in a lower after-hours communications
expectation climate, they are free to use ICTs in the work-to-family direction as they choose.
While segmenters may elect not to use work-to-family ICTs at all or only for very specific
purposes, this use may benefit integrators who gain from it desired flexibility. Thus in high afterhours communications expectation climates, work-to-family ICT use is less flexible, functions as
a job demand, and should be more exhausting. In low after-hours communications expectation
climates, work-to-family ICTs can function as a job resource, at least for integrators. The net
effect of work-to-family ICT use on exhaustion should therefore be negative in lower after-hours
communications expectation climates. In other words, using ICTs is more demanding, on
average, when after-hours communications expectations are high than when they are low. This

39

moderation is shown in Figure 1 with the arrow linking work-to-family ICT use to emotional
exhaustion and the arrow pointing to it from after-hours communications expectation.
Hypothesis 3: After-hours electronic communications expectations moderate the relationship
between work-to-family ICT use and emotional exhaustion such that it is positive when afterhours communications expectations are high and negative when after-hours communications
expectations are low.
While after-hours communications expectations may be a factor in how employees
experience the use of ICTs, they might also play a role in how employees experience their own
segmentation preferences regardless of their actual ICT use. Conceptually, there is little reason to
expect that segmentation preferences alone are associated with emotional exhaustion; neither
segmenting nor integrating is inherently better or worse for employee well-being. Research
suggests that it is the fit between segmentation preferences and actual ability to segment that
drive individual outcomes (Edwards & Rothbard, 1999; Kreiner, 2006). Once again, though,
after-hours communications expectations might be a contextual factor that changes how an
individual’s preferences are experienced.
For example, being in a high after-hours communications expectation climate may be
especially demanding on segmenters, who prefer to keep their work role demands out of the
family domain. Given an identical level of after-hours communications expectations, a
segmenter may perceive higher levels of work pressure than an integrator, even if he or she is not
using ICTs very frequently. Work pressure is one type of well-recognized job demand (R. T. Lee
& Ashforth, 1996). Hypothesis 2 suggests that segmenters will be more likely to use work-tofamily ICTs in higher after-hours communications expectation climates, but regardless of their

40

levels of ICT use, segmenters may experience the pressure to stay connected differently than
integrators, i.e. as a job demand.
For integrators, the pressure to stay connected is compatible with their own preferred
boundary management preferences and so this pressure should be less demanding. For
segmenters, after-hours communications expectations are incompatible with their preferences,
creating dissonance between actual and desired states of boundary management, resulting in
increased stress. Thus, while integrators may not be affected by after-hours communications
expectations, simply being in an environment of high after-hours communications expectations
may create stress for segmenters regardless of their actual ICT use. This is shown in Figure 2.1
with the arrow linking work-to-family segmentation preferences to emotional exhaustion
(through work-family conflict as described below) and the arrow pointing to it from after-hours
communications expectations.
Hypothesis 4: After-hours electronic communications expectations moderate the relationship
between work-to-family segmentation preferences and emotional exhaustion such that it is
positive when after-hours communications expectations are high.
Hypotheses 3 and 4 link work-to-family ICT use and work-to-family segmentation
preferences to emotional exhaustion, respectively, moderated by after-hours communications
expectations. However, these different pathways to emotional exhaustion may not operate the
same way; they may be related to emotional exhaustion through different explanatory
mechanisms. In other words, after-hours communications expectations may be creating different
types of job demands in each interaction. Identifying and testing these mechanisms may shed
additional insight as to why these variables might be related to levels of emotional exhaustion.
Lee and Ashforth (1996) identify eight job demands, some of which can be used to show

41

mechanisms through which the hypothesized relationships may be functioning. For the current
study, work pressure and role conflict are most relevant.
Work pressure is one type of job demand strongly associated with emotional exhaustion
(Bakker, Demerouti, & Dollard, 2008; Demerouti, Bakker, & Bulters, 2004; Demerouti et al.,
2001). When workers experience high workload or time pressure, they experience a loss of
psychological and physical resources which can lead to emotional exhaustion (Demerouti et al.,
2001). Role conflict is a second relevant job demand. Work-family role conflict occurs when
work role demands prevent a person from completing demands associated with the family role
(Frone, Russell, & Cooper, 1992; Greenhaus & Beutell, 1985). These conflicts are most typically
time-based or strain-based. In time-based conflict, the amount of time needed for work role
demands takes away from time needed to adequately meet family role demands. In strain-based
conflict, strain from work role demands makes it more difficult to meet family role demands.
Hypothesis 3 states that that work-to-family ICT use will be more positively related to
emotional exhaustion under high after-hours communications expectation climates. I argue
above that under such conditions, ICT use is a result of organizational technical control. Afterhours communications expectations are conceptualized as a measure of a job demand that could
be categorized as work pressure as they are an expectation for additional work to be performed.
Lee and Ashforth’s (1996) meta-analysis showed a significant, positive relationship between
work pressure and emotional exhaustion.
When considering work-to-family segmentation preferences, after-hours communications
expectations may function as more than work pressure; they may also be a source of conflict
between work and family roles. Segmenters in high after-hours communications expectation
climates may experience greater work-family role conflict in addition to work pressure because

42

after-hours communications expectations are incompatible with their preferred role boundary
management preferences. After-hours communications expectations compel segmenters to
integrate work demands into the family domain against their will. This creates the perception that
desired family role time is being taken up by work demands. Similarly, it increases the likelihood
that stress from work demands will permeate into the family domain because work tasks are
being performed in the home. Thus, segmenters will experience after-hours communications
expectations as increased role conflict while integrators will not.
Research also suggests that work-family role conflict and perceptions of work pressure
may exacerbate one another (Demerouti et al., 2004). Integrators may experience less role
conflict from after-hours communications expectations because the pressure to use ICTs to work
from home is compatible with their preferred role boundary management strategy. Whether
segmenter or integrator, though, an individual should still experience more work pressure from
higher after-hours communications expectations. Thus work-family conflict should only partially
mediate the after-hours communications expectation and segmentation preference interaction on
emotional exhaustion.
Hypothesis 5: The interaction of after-hours electronic communications expectations and workto-family segmentation preferences on emotional exhaustion is partially mediated by workfamily conflict.
Finally, stress has important consequences for employees’ working lives. Over time,
stress takes its toll and employees may seek less stressful employment elsewhere. Research on
job demands and resources has theoretically linked emotional exhaustion to increased turnover
intentions (e.g., Schaufeli & Bakker, 2004) and this link has received wide empirical support
(e.g., Lee & Ashforth, 1996). Studies also suggest that inability to manage ones role boundaries

43

effectively contributes to turnover, particularly for teleworkers, suggesting an overall link
between work-family role boundary management and turnover (Kossek et al., 2009, 2012;
Lautsch, Kossek, & Eaton, 2009). Consistent with this line of research, emotional exhaustion
should be positively related to turnover intentions. This is shown in Figure 2.1 by the arrow
linking emotional exhaustion to turnover intentions.
Hypothesis 6: Emotional exhaustion is positively related to turnover intentions.
Together these hypotheses suggest a complex set of interactions between an employee’s
preferred boundary management style, pressure to manage boundaries in a certain way, and
actual use of information and communicative technologies to manage boundaries in that way. At
the core of this relationship is whether the organization pressures employees to use ICTs to
integrate work into the family domain. In other words, an employee’s level and experience of
ICT use will depend on whether the organization pushes employees to use ICTs to intensify
work or allows them to use ICTs to enhance their work-family role boundary control.
2.5

Methods

2.5.1 Measures
The complete final survey instrument, omitting extraneous measures unused in the
analysis, can be found in Appendix A. Numerous demographic and work characteristics were
assessed in the survey: age, number of children under 18 living in the household, marital status,
and whether the employee was involved in caring for an elderly family member. Participants also
reported whether they worked in the human resources field, the industry they worked in, and
weekly hours worked both at the workplace and at home. With the exception of working in the
human resources, which was used as a screen for inclusion in the sample, these measures
functioned as control variables.
44

The purpose of including these variables as controls is to account for their potential
influence on the dependent variables and demonstrate that the focal independent variables (afterhours communications expectations and work-family ICT use) are related to the dependent
variables even when partialling out their relationships with these other variables (Carlson & Wu,
2012). Distribution of working time at home versus the workplace, being in a non-manufacturing
industry, and family demands are plausibly positively related to the degree to which an employee
uses technology for work-family purposes and their likelihood of being in an organization with
higher after-hours communications expectations. Older employees and women are more likely to
have family demands which may result in greater baseline levels of emotional exhaustion or time
spent working from home. Given the frequency of employees in the sample that fit in these
categories omission of these variables could potentially result in spuriously high relationships
between the independent and dependent variables. Similarly, these variables are plausibly related
to the amount of emotional exhaustion and work-family conflict experienced by employees.
Several measures assessed individuals’ relationship with technology. After-Hours
electronic communications expectations (α=0.86) describe the degree to which the organization
expects to be able to reach the employee through communicative technology while he or she is
away from the workplace. This seven-item measure was adapted from Fender (2010). Though it
showed good psychometric properties in that study, some items were slightly reworded to
eliminate possible ambiguities and to more broadly apply to various mediums of information
communicative technology.
ICT use for home and work purposes were measured using two scales developed and
validated for this study, as described in the following pilot data analysis section. Several items
for the scale were based on items developed for a similar scale by Fender (2010), titled "response

45

to electronic communications behaviors." However, this scale does not differentiate between
technology use for work and family purposes or the location of the technology use and therefore
required adaptation in the current study. The work-to-family ICT use (WF ICT use, α=0.94) and
family-to-work ICT use (FW ICT use, α=0.94) scales used here describe the extent to which
employees use information communicative technology to meet work demands while at home and
family demands while at work, respectively.
FW ICT use was included as a control for a more general tendency to use technology
(potentially confounding the measurement of WF ICT use) and to ensure that analyses captured
attitudes and behaviors about work-to-family role management separately from those associated
with family-to-work role management, as research suggests important differences between the
two (Frone et al., 1992; Kossek et al., 2012). This is important as individuals may have a general
propensity toward technology use regardless of work-to-family or family-to-work direction and
individuals vary substantially in their ability and willingness to use technology (e.g., King & He,
2006).
Several scales measured the state of employees’ work-family role management. Workfamily conflict (α=0.93) was measured using a five item widely used and well-accepted scale
(Netemeyer, Boles, & McMurrian, 1996) and assessed the degree to which employees' work
demands prevented them from meeting family demands. Work-family segmentation preferences
(α=0.90) were measured using a scale from Kreiner (2006) and assessed the degree to which
employees preferred to keep their working roles in the work domain. The measure was adapted
to form a Family-work segmentation preferences (α=0.85) scale, used as a control for reasons
similar to those for including the FW ICT use scale, i.e. to ensure the measure was capturing the
appropriate direction of the preference rather than a general propensity to segment.

46

Work role identity (α=0.80) and family role identity (α=0.93) were measured using two
items each, from Kossek, Lautsch, and Eaton (2006). These items measured the strength with
which a person identifies as work or family-focused and were also included as controls as they
may influence an individual’s perceptions of role conflict and stress. In addition, work methods
autonomy (α=0.88) was included as a control using the three-item scale from the Work Design
Questionnaire (Morgeson & Humphrey, 2006), a carefully validated measure covering many
specific facets of work design. This was included as those with higher autonomy are more likely
to be able to use ICTs to perform work more flexibly, and the study seeks to show the
importance of WF ICT use and after-hours communications expectations above and beyond job
design.
Finally, two dependent variables were measured. Emotional exhaustion (α=0.92) was
assessed using a subscale of the Maslach Burnout Inventory (Maslach, Jackson, & Leiter, 1997).
This scale defines emotional exhaustion as "a state of exhaustion in which one is cynical about
the value of one's occupation and doubtful of one's ability to perform." The MBI has been used
in over 70 studies as early as 1996 (R. T. Lee & Ashforth, 1996). This measure was chosen not
only for its general acceptance in the literature but also for its careful development. Turnover
intentions (α=0.94) were measured using a straightforward two-item measure adapted from
Boroff and Lewin (1997) and assessing the employee’s plans to voluntarily quit within the next
year. Prior to conducting primary research, I conducted a pilot study to provide empirical
validation and support for the scale adaptations and development described above.
2.5.2 Pilot Survey Analysis
In order to provide evidence for the validity and reliability of some new and adjusted
measures as well as to test the survey instrument for errors and completion time, I conducted a
47

small pilot of the survey. Participants were students in a human resource management master’s
degree program. Students were invited to participate via e-mail, and data was collected online
from 51 students. The full survey took participants typically between 10 to 15 minutes.
2.5.2.1 Scale Reliability
Reliability was acceptable for all scales except for turnover intentions, however given the
nature of the sample the turnover intentions measure is likely not appropriate. Many students of
were likely not employed or knew they would be quitting their jobs upon graduation. Despite
this, reliability was very close to practical acceptable levels and thus the scale was retained for
the final survey (in which an acceptable reliability coefficient was indeed observed). Alphas for
each scale from the pilot sample are presented in Table 2.1.
2.5.2.2 ICT Use Scale Construction
Work-to-family (WF) and family-to-work (FW) ICT use were scales constructed for the
current study. These scales are heavily based on the Receptive Electronic Communications scale
used by Fender (2010). The primary differences are that Fender’s items were specific to smart
phones and cell phones while the present items address technology more generally, with the
intent of including other common ICT mediums such as laptop computers. As a result some
items were modified or rewritten. Additionally, Fender’s scale was work-to-family only; the
present study adapts the work-to-family items into a separate family-to-work scale by changing
the direction of the item wording to use as a control for more general technology use.
Additionally, I sought to reduce the length of the scale in order to minimize the burden on
participants.

48

Table 2.1: Number of scale items and alpha coefficients for the pilot study.
Scale
Items
Alpha coefficient
1
13
0.956
Work-to-family Technology Use
1
13
0.958
Family-to-work Technology Use
Work Methods Autonomy
Emotional Exhaustion
Turnover Intentions
Work Identity
Family Identity
Work-Family Separation Preference
Family-Work Separation Preference
Work-Family-Conflict
After-hours Electronic Communication Expectations
2
Work-to-family Technology Use

3
3
2
2
2
4
4
5
7
4

0.916
0.933
0.685
0.819
0.908
0.885
0.884
0.943
0.934
0.941

Work-to-Family Technology Use

4

0.944

2

1

Notes: Scales containing excess items for validation purposes.
2

Final version scales.

In order to reduce the overall number of scale items, remove lower quality items, and
verify that WF and FW tech use were indeed separate constructs (rather than representing for
example one more general technology use construct), I performed an exploratory factor analysis
using all 26 initial scale items. In accordance with best practices (Costello & Osborne, 2005), I
used a factor analysis rather than a principal components analysis. Whereas principal
components analysis assumes no measurement error, factor analysis removes error variance and
subsequently corrects item correlations. Given the relative normality of the data, I used
maximum likelihood as the extraction method, again consistent with best practices (Fabrigar,
Wegener, MacCallum, & Strahan, 1999). I used a scree plot to determine the number of factors
to retain, in which the researcher keeps components above the inflection point of the eigenvalue
plot. Finally, I used a direct oblimin oblique rotation to allow the factors to correlate as they are
likely partially related (e.g., via a general propensity to use technology).

49

The scree plot is given in Figure 2.2 and revealed two factors as anticipated, explaining
47.67% and 19.91% of the variance, respectively. These factors were easily interpreted as WF
and FW ICT use. The factor structure matrix can be found in Table 2.2 and includes four factors
as the third and fourth yielded eigenvalues greater than one. However, the scree plot along with
theoretical guidance used in the conceptualization of the scales suggests that these factors are not
useful and should not be retained. Other models using different rotation and extraction methods
conducted as a robustness check yielded similar results and are not presented. Based on these
factor structures and a need for parsimony, four items were retained for each scale which had the
most consistent factor loadings. For WF technology use, WF items 1, 3, 5, and 7 were retained.
For FW technology use, FW items 2, 4, 7, and 9 were retained. These items are presented in
Table 2.3. Alpha coefficients of these reduced-item scales were nearly identical to their 13-item
counterparts and are also presented in Table 2.1.
2.5.3 Full Survey Analysis
2.5.3.1 Sample and Design
Based on the satisfactory results of the pilot data collection, I continued to primary data
collection having made the changes in the survey noted above and with the additional change
that the survey was split into two waves to help alleviate potential common method bias
(Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Participants were alumni of a human
resources masters degree program and were invited to complete the survey via e-mail. Human
resource managers are an ideal sample because their job requires frequent communication with
other employees and tasks in the human resource management function are largely knowledgebased. Thus, the job functions of human resource managers are amenable to after-hours work via
e-mail and cell phone communication because their institutional role is enacted through
50

knowledge work and communication (Kossek, Dass, & DeMarr, 1994). Based on the size of the
email list, a reasonable but conservative estimate of the response rate is approximately 15%.
However, it is impossible to know whether all e-mail addresses used are current and actively
monitored. Though this response rate is a bit low, at about one standard deviation below the
mean for web-based surveys, response rate is less important when representativeness is less
important such as in this case when the sample was chosen primarily based on occupation (C.
Cook, Heath, & Thompson, 2000).
Figure 2.2: Scree plot for work-family and family-work technology use items.

51

Table 2.2: Work-family and family-work technology use factor structure matrix.
Factor
Item
1
2
3
4
wftech1
.872
.292
-.152
.453
wftech2
.839
.327
.078
.505
wftech3
.936
.369
.009
.323
wftech4
.869
.350
.047
.564
wftech5
.888
.312
-.208
.722
wftech6
.823
.389
-.042
.737
wftech7
.890
.252
-.197
.457
wftech8
.741
.354
.268
.445
wftech9
.799
.313
.026
.500
wftech10
.412
.209
.068
.717
wftech11
.726
.505
-.159
.636
wftech12
.654
.338
-.024
.715
wftech13
.713
.299
-.049
.641
fwtech1
.326
.760
-.465
.034
fwtech2
.197
.867
-.170
.134
fwtech3
.319
.813
-.055
.321
fwtech4
.364
.900
-.320
.290
fwtech5
.138
.731
-.773
.069
fwtech6
.167
.597
-.750
-.007
fwtech7
.277
.874
-.307
.097
fwtech8
.373
.867
-.128
.251
fwtech9
.356
.918
-.243
.226
fwtech10
.305
.764
-.357
.170
fwtech11
.524
.668
-.115
.314
fwtech12
.197
.585
-.298
.120
fwtech13
.402
.805
-.328
.399
Notes: Extraction Method: Maximum Likelihood.
Rotation Method: Oblimin with Kaiser normalization.

52

Table 2.3: Retained work-family and family-work technology use items
Item
Item Text
In a typical month, to what extent do you…
WF 1
...use technology to keep up with work matters while at
home?
WF 2
...use technology to communicate with your workplace
outside of normal work hours?
WF 3
...use technology to perform work tasks while at home?
WF 4
...respond to someone from work using technology to
communicate with you when you're at home?
WF 5
...use technology to stay connected to your work no matter
where you are?
WF 6
...use technology to stay connected to your work no matter
what time it is?
WF 7
...use technology to access information pertaining to work
tasks while at home?
WF 8
...experience interruptions at home from work-related
communications?
WF 9
...provide status reports after working hours via technology?
WF 10
...check for technology-based communications as soon as you
got up in the morning?
WF 11
...use technology to contact others in the organization to work
on problems you learned about after typical working hours?
WF 12
...actively check for messages using technology after typical
working hours?
WF 13
...use technology to immediately respond to work-related
communications after typical working hours?
...use technology to keep up with family matters while at
FW 1
work?
FW 2
...use technology to communicate with your family or friends
during normal work hours?
FW 3
...use technology to perform personal tasks while at work?
FW 4
...respond to a family member using technology to
communicate with you when you're working?
FW 5
...use technology to stay connected to your family no matter
where you are?
FW 6
...use technology to stay connected to your family no matter
what time it is?
FW 7
...use technology to access information pertaining to your
family while working?
FW 8
...experience interruptions while working from family-related
communications?
FW 9
...check in with family during working hours via technology?
FW 10

...check for technology-based communications from family as
soon as you get to work?
53

Retained?
Yes

Yes
Yes

Yes

Yes
Yes

Yes

Yes

Table 2.3 (cont'd)
Item
FW 11
FW 12
FW 13

Item Text
In a typical month, to what extent do you…
...use technology to address family-related problems during
the workday?
...actively check for messages from family using technology
after typical working hours?
...use technology to immediately respond to family-related
communications during typical working hours?

Retained?

Human resource managers were chosen because of their status as having a heavily
knowledge-based occupation with primary tasks amenable to ICT use as well as the availability
of participants in this occupation. Based on this sample, results should be generalizable to other
occupations with a core function based around knowledge management and a similar capacity
for work to be performed or enhanced by the use of ICTs. Other occupations to which this
research might be directly applicable would be general managers, software developers, or those
in marketing or public relations. The degree these results are applicable and important to an
employee in a particular occupation are contingent on that occupation’s tasks being performable
via ICTs. Access to human resource managers was available through personal networks thus
making the sample desirable from both a research design and data collection perspective.
Additionally, this sample provided access to human resource managers across a variety of
organizations, allowing for necessary variation in the key independent variable, after-hours
communications expectations, which is based on organization-level perceptions.
In the primary sample, the survey was conducted in two waves in order to combat
potential common method bias (Podsakoff et al., 2003). In addition to temporal separation of the
measurement of the independent and dependent variables, I also ensured anonymity of responses
and used measures with a variety of scale endpoints consistent with recommendations to
minimize common method bias (Podsakoff et al., 2003). Wave 2, administered approximately
54

one month after Wave 1, contained scales for work-family conflict, emotional exhaustion, and
turnover intentions while all others, described in detail below, appeared in Wave 1.
Wave 1 yielded 233 total respondents who were then invited to participate in Wave 2. Of
these 233, 173 both participated in Wave 2 and could be matched to their Wave 1 data, resulting
in an approximately 26% attrition rate. Pairwise correlations of all study variables reveal that
Wave 2 participants were slightly less likely to work in the manufacturing industry. Finally, 30
participants were not employed in human resource management occupations. These 30 were
dropped from the study as they may not be involved in knowledge-based work amenable to ICT
use. The final sample size was 203, with 157 participants providing full data and 46 providing
data for only Wave 1.
The final sample had an average age of 41. Of the participants, 53% were female, 68%
were married, 71% had children under 18 at home, and 6% had elder care responsibilities.
Participants worked on average 41.86 hours per week at the workplace and 10.3 hours per week
at home. Finally, 43% worked in manufacturing and 30% worked in service industries. These
demographics served as controls in all analyses, with “other” being the omitted category for
industry.
2.5.3.2 Analytic Techniques
The study hypotheses were assessed using six single-stage ordinary least squares
regression models. To establish mediation as described in Hypothesis 5, I followed Baron and
Kenny’s (1986) four step procedure as well as Preacher and Hayes’ (2004) direct test of the
indirect effect, described below. In order to establish mediation as defined by R. Baron and
Kenny (1986), four conditions must be met. First, the independent variable must be shown to be
related to the mediator. Second, the independent variable must also be shown to be related to the
55

dependent variable. Third, the mediator must be shown to be related to the dependent variable.
Finally, the relationship established in the second requirement must be reduced (to zero in the
case of full mediation) upon introduction of the mediator into the equation.
Moderation, on the other hand, occurs when the strength of the relationship of two
variables is dependent on a third (R. Baron & Kenny, 1986; Preacher, Rucker, & Hayes, 2007).
Moderation is shown in the present study by including the product of the interacting variables in
the model. The nature of these moderation effects is more clearly shown through simple slope
plots. I calculate and plot simple slopes using techniques described by J. Cohen, Cohen, West,
and Aiken (2003). Simple slopes are estimates of the relationship between the independent and
dependent variable at one standard deviation below, one standard deviation above, and equal to
the mean of the moderating variable and aid with the interpretation of moderation effects (J.
Cohen et al., 2003). I also calculate the standard error of the slope at each of the three levels of
the moderating variable and the resulting sample T statistic for difficult to interpret interaction
effects.
Mediated moderation occurs when a moderation effect is mediated by another variable
(Muller, Judd, & Yzerbyt, 2005). In order to establish mediated moderation, one shows that two
variables have an interaction effect on a third dependent variable, and that this interaction is
mediated by a fourth mediating variable (R. Baron & Kenny, 1986; Preacher et al., 2007).
Mediation is typically assessed using the same Baron and Kenny four-step procedure outlined
above, but using the interaction term as the key independent variable to be mediated.
An increasingly common critique of the Baron and Kenny (1986) mediation rules are that
they do not require an empirical test of the total indirect effect. Preacher and Hayes (2004)
provide reasons for the importance of this test. First, they explain that the coefficient of the

56

independent variable may become statistically insignificant with a very small reduction in
magnitude upon the introduction of the mediating variable. Similarly, a practically large
coefficient change may still be statistically significant, suggesting a lack of mediation. Second,
they suggest that testing the difference between the direct and indirect effects is a more
straightforward way of testing mediation.
One way to test the indirect effects is with the Sobel test, but the Sobel test assumes a
normal distribution of the difference between the direct and indirect effects of the independent
variable on the dependent variable. In fact, this difference is often not even symmetrical but
rather positively skewed and kurtotic (Bollen & Stine, 1990; Lockwood & MacKinnon, 1998;
MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002; Preacher & Hayes, 2004; Preacher et
al., 2007; Shrout & Bolger, 2002). Instead, Preacher and Hayes (2004) suggest using
bootstrapping. Because indirect effects approach normality in larger samples, bootstrapping is
even more strongly recommended for studies with smaller samples (Preacher et al., 2007).
Bootstrapping makes no assumptions about the sampling distribution of the indirect
effects. It is a technique which takes a designated number of samples from the study sample,
treating the study sample as its own population from which confidence intervals are constructed.
Bootstrapping is therefore a resampling technique. The percentile-based confidence intervals
generated by bootstrapping can be further adjusted to correct for bias (MacKinnon, Lockwood,
& Williams, 2004; Preacher et al., 2007). The only major drawbacks of bootstrapping are
computational power requirements and the fact that confidence intervals may change when
bootstrapping is applied again as it uses random resamples (Preacher et al., 2007). These
drawbacks are minor as modern computers have more than sufficient computational power and a
large number of resamples provides a more stable confidence interval. To assess the indirect

57

effect posited in Hypothesis 5, I use bootstrapping with bias-corrected confidence intervals in
addition to the more typical Baron and Kenny (1986) four step procedure. I use a relatively large
number of resamples (5,000) to minimize differences in confidence intervals due to resampling.
A potential problem in the present study is common method bias (also known as common
method variance). Common method bias occurs when participants’ responses are related due to
similarities in the method in which they are collected rather than because of an underlying
theoretical relationship (Podsakoff et al., 2003). Typically, this occurs in survey designs in which
all measures are similarly structured. For example, the respondent’s mood may impact the way
they answer items at a particular time, creating an undesirable correlation among psychological
measurements (Podsakoff et al., 2003). It remains unclear whether common method bias is a
substantial problem in organizational research as some argue its effect is typically not obtrusive
(Lance, Dawson, Birkelbach, & Hoffman, 2010; Spector, 1987).
Still, I took numerous methodological precautions in the survey design, described above.
In addition to techniques used in survey design to minimize common method bias, there also
exist some techniques to test for the presence of common method bias within the data. These
tests do not necessarily detect common method bias, however if they are failed then it is likely
that common method bias is present and a significant source of variation in the measures. I
conducted a Harman’s Single Factor test, which is frequently used to assess the extent of
common method bias in a survey (Podsakoff et al., 2003). This test was conducted using
exploratory factor analysis and items from all items collected in the survey, including those not
used in the current study (17 scales in total). The test yielded 40 factors, 13 of which had an
eigenvalue greater than 1 with the most variance explained by a single factor being 19.8%. These

58

results are consistent with a lack of substantial common method bias. While they do not
guarantee that it is not present, this test suggests that it is not problematic.
2.6

Results
Measure intercorrelations, reliabilities, and descriptive statistics can be found in Table

2.4. Correlations indicate that several of the demographic controls were related to WF ICT use as
expected, providing some justification for their inclusion in the regression models (Carlson &
Wu, 2012). WF ICT use is higher among older employees, married employees, and those with
children. It is also higher among those who report more FW ICT use, more autonomy, and higher
work identity. Women report higher work-family conflict while older employees reported less
emotional exhaustion. Work identity is also positively related to higher after-hours
communications expectations and work-family conflict. FW segmentation preferences are
positively related to WF segmentation preferences. Home work hours are unsurprisingly
positively related to after-hours communications expectations and WF ICT use.
Table 2.5 shows the results of the regression models. Hypothesis 1 stated that
segmentation preferences would be negatively related to work-family ICT use. Model 1 indicates
support for Hypothesis 1, as shown by the negative coefficient on the work-family segmentation
preferences variable (β=-0.642; p<0.01). Furthermore, in support of Hypothesis 2, Model 1 also
shows this negative coefficient becomes more positive at higher levels of after-hours electronic
communication expectations. This is shown by the positive and significant coefficient on the
interaction of after-hours communications expectations and work-to-family segmentation
preferences (β=0.152; p<0.05). Thus, Model 1 shows that integrators are more likely to use ICTs
for work while at home than segmenters, but that this difference is smaller when after-hours
communications expectations are higher.
59

Hypothesis 3 stated that work-family ICT use would be positively related to emotional
exhaustion under higher after-hours communications expectations, but negatively related to
emotional exhaustion under lower after-hours communications expectations. In support of
Hypothesis 3, Model 2 in Table 2.5 shows that the interaction of work-to-family ICT use and
after-hours communications expectations was statistically significant and positive (β=0.316;
p<0.05). Figure 2.3 provides a graphical representation of the simple slopes of the interaction so
the nature of the effect can be more clearly interpreted. Table 2.6 shows the results of the simple
slope analysis based on J. Cohen et al. (2003) and described above. In opposition to Hypothesis
3, at high levels of after-hours communications expectation, work-family ICT use was unrelated
to emotional exhaustion as the slope was not statistically significantly different from zero
(T=0.62; p>0.05). On the other hand, in support of Hypothesis 3, at low levels of after-hours
communications expectations, greater work-family ICT use was associated with lower emotional
exhaustion (T=-2.28; p<0.05). Overall, Hypothesis 3 was partially supported. Surprisingly,
Model 2 also revealed an unhypothesized significant and negative main effect of after-hours
communications expectations on emotional exhaustion (β=-1.813; p<0.05). This may be a result
of controlling for multiple interactions involving after-hours communications expectations, and
this finding is addressed in greater detail in the discussion and post-hoc analysis.
Hypothesis 4 stated that segmentation preferences are more positively related to
emotional exhaustion when after-hours communications expectations are higher. Model 2 also
tests this interaction and finds that it is statistically significant and positive, in support of
Hypothesis 4 (β=0.273; p<0.05). Hypothesis 5 states that this interaction is partially mediated by
work-family conflict. Models 3, 4, and 5 in Table 2.5 together test Hypothesis 5. According to
Baron and Kenny (1986), several steps are necessary to establish mediation. The first step,

60

showing a relationship between the mediated variable and the dependent variable, is met by the
support of Hypothesis 4 in Model 2, which shows that the interaction is related to emotional
exhaustion. The second step is to show that the independent variable is related to the mediator.
Model 3 regresses the interaction on work-family conflict, showing a positive and significant
relationship (β=0.285; p<0.01), meeting the requirements of step two.
The nature of this interaction effect is shown graphically in Figure 2.4. Table 2.7 shows
the tests of the associated simple slopes. The relationship between work-family segmentation
preferences and work-family conflict is statistically significantly positive at high levels of afterhours communications expectations (T=3.04; p<0.01). The relationship is only marginally
significant, but negative, at low levels of after-hours communications expectations (T=-1.54;
p<0.10). This suggests that integrators experience little work-family conflict attributable to high
after-hours communications expectations, but segmenters experience higher levels of workfamily conflict under higher after-hours communications expectations. Additionally, in low afterhours communications expectations, there is a non-significant trend associating segmentation
with lower work-family conflict.
Returning to the mediation of Hypothesis 5, the third step is to show that the mediating
variable is related to the dependent variable without the independent variables present. Table 2.5,
Model 4 shows that work-family conflict, without the presence of any interaction effects, is
positively related to emotional exhaustion (β=0.488; p<0.01). This meets the criteria of step
three. The fourth and final step of the mediation test is to show that the relationship between the
independent variable (once again, the interaction of after-hours communications expectations
and work-to-family segmentation preferences) and the dependent variable (emotional
exhaustion) is reduced upon inclusion of the mediating variable. This relationship was reduced to

61

non-significance in Model 5 (β=0.140; p>0.05), consistent with full rather than partial mediation.
Thus Hypothesis 5 was partially supported; the interaction of after-hours communications
expectations and work-to-family segmentation preference on emotional exhaustion was fully
mediated by work-family conflict. Segmenters experienced higher work-family conflict under
higher after-hours communications expectations, and this fully accounted for the similar
relationship between segmentation preferences, after-hours communications expectations, and
emotional exhaustion.
To confirm this effect, as discussed above in the analytical strategy section, it is
important to test the statistical significance of the indirect relationship between the independent
variable and the dependent variable (Preacher & Hayes, 2004). The results of this test are shown
in Table 2.8. If the confidence interval does not include zero, then the indirect effect is
statistically significant at the 5% level. Bias-uncorrected bootstrapped standard errors are
provided for comparison, and the uncorrected normal-based confidence interval does include
zero. The more accurate percentile, bias-corrected, and accelerated bias-corrected confidence
intervals on the other hand do not include zero, providing further support for the mediation
described in Hypothesis 5.
Finally, Hypothesis 6 predicted that emotional exhaustion would be positively related to
turnover intentions. Controlling for work-family conflict as well as all other previously
mentioned relationships, Model 6 in Table 2.5 shows that the coefficient for the hypothesized
relationship is positive and statistically significant (β=0.342; p<0.01), supporting Hypothesis 6.
Figure 2.5 below shows the theoretical model presented in Figure 2.1, but revised to be
consistent with the empirical relationships supported in the present study. Specifically, the partial
mediation proposed in Hypothesis 5 is represented as full rather than partial mediation.

62

Table 2.4: Measure intercorrelations, alphas, means, and standard deviations
1
2
3
4
5
6
1. Industry: Manufacturing
*
-0.54
2. Industry: Service
3. Work hours: home

-0.16

4. Work hours: work

0.15

5. Age
6.Married

0.17

*

*

-0.17
*
0.12
-0.22
-0.11
0.08

9. Female

-0.04
-0.05
0.12

10. WF ICT use

0.00

11. FW ICT use

-0.31
-0.06

0.29
0.14

13. Work identity

0.09

-0.13

14. Family identity

-0.13

0.05

15. WF segmentation pref.

0.09

-0.12

16. FW segmentation pref.

-0.03

0.02

17. AECE

0.06

18. Emotional exhaustion

0.06

19. Turnover Intentions
20. WFC
Mean
Sd

7. Kids
8. Elder Care

12. Autonomy

*

*

9

10

11

0.42
0.09
-0.16

0.03
0.02

-0.03

-

*

*

0.01

-0.06

0.92

*

0.15

0.01

0.21

12

-

-0.67
0.10

-0.10

-

0.11

-0.13

0.43

0.06
0.08
0.12

-0.05

*

-0.02

8

-

-0.03
0.05
-0.09
*

7

*
*

-0.11
-0.14

0.23
0.07

-0.20

-0.38

*
*

*

*

0.47
0.09

-0.18
-0.14

0.21
0.11

*

*

*

0.22

*

0.19
0.06

-0.19
0.14

*

-0.27
*
0.12
-0.24
-0.02
0.06
*

0.07

-0.16

0.20
0.07

-0.06

-0.12

0.09

0.09
0.43
0.50

-0.11
0.30
0.46

0.12
10.30
10.94

0.12

*

0.23
0.02

0.26
0.19

*

0.16
0.07

0.15

0.16

0.34
-0.10

0.32
-0.10

-0.09

0.01

0.00

-0.06

0.14

0.15

-0.07

*

-0.19
0.03
0.19

*
*

0.10
*

*

0.04

-0.35
-0.09

-0.22
0.05

0.09
41.86
13.45

-0.02
41.46
13.08

0.10
0.68
0.47

63

*
*

0.17
0.05

-0.03

-0.02

0.35

-0.15

-0.01

*

0.03

-0.05

0.38
0.09

*

*

0.20
0.06

*

0.88
0.22

*
*

0.22
0.21
*
*
0.00 -0.35 -0.10 -0.23
*
0.05 -0.06 -0.01
-0.20
*
-0.01 0.54
-0.05 0.10
*

-0.04 -0.16
-0.02 -0.03

0.25
0.05

0.08
0.71
1.01

0.21
0.53
0.50

-0.08
0.06
0.24

*

0.95

*

-0.06
-0.02
*

0.20
4.12
1.01

*

-0.07 -0.24
*
-0.05 -0.19
-0.08
3.39
1.13

-0.06
4.15
0.76

Table 2.4 (cont’d)
1. Industry: Manufacturing
2. Industry: Service
3. Work hours: home
4. Work hours: work
5. Age
6.Married
7. Kids
8. Elder Care
9. Female
10. WF ICT use
11. FW ICT use
12. Autonomy
13. Work identity
14. Family identity
15. WF segmentation pref.
16. FW segmentation pref.

13

14

15

0.80
0.07
0.00
0.11

0.93
0.04
-0.09

0.90

*

18. Emotional exhaustion

0.23
0.05

19. Turnover Intentions

0.04

20. WFC
Mean
Standard deviation

0.19
4.40
0.61

17. AECE

*

*

-0.04
*

-0.21
0.07

*

-0.26
3.61
1.02

0.34

16

18

19

20

*

0.85

*

0.11

*

0.12 0.18
0.92
*
*
0.11 0.19 0.48
0.94
*
*
*
0.03 0.32 0.45 0.27 0.93
3.32 2.76 3.87 2.30 2.77
0.86 0.85 1.45 1.31 1.01

-0.17
0.30

17

*

0.24
0.05
3.49
0.98

0.86
*

Notes: p<0.05
Alphas on diagonal
AECE=After-hours electronic communications expectations
64

Table 2.5: Regression models
Model 1
Dependent
WF ICT Use
variable:
B
SE
Manufacturing 0.047
0.143
Service
-0.311
0.154
Work hours at
0.029
0.008
home
Work hours at
0.000
0.007
work
Age
-0.002
0.006
Married
Kids
Elder care
Female
Work-family
ICT use
Family-work
ICT use
Autonomy
Work identity

0.162
0.033
-0.132
-0.167
0.158
0.170

0.153
0.069
0.243
0.131
-

Model 2
Emotional
Exhaustion
B
SE
-0.434
0.262
*
0.286
-0.586
0.048

**

0.025
-0.027
-0.144
-0.003
-0.497
0.397
-0.960

*

**

Model 3
B
-0.006
-0.153

SE
0.188
0.203

Model 4
Emotional
Exhaustion
B
SE
-0.329
0.252
-0.470
0.276

0.016

0.018

0.011

0.040

0.016 0.040

0.013

0.011

0.009

0.020

0.012 0.020

0.011

0.007

0.008

-0.028

**

0.204
0.092
0.313
0.171

-0.266
0.062
-0.285
0.212

0.010 -0.030
0.279 -0.359
0.122 0.025
0.422 -0.382
0.236 0.204

**

WF Conflict

*

Model 5
Emotional
Exhaustion
B
SE
-0.407
0.252
-0.503
0.273
*

0.284
0.129
0.441
0.240

0.512
-0.060
-0.246

0.341

-0.056

0.242

-0.257

0.154 -0.930

0.435

*

Model 6
Turnover
Intentions
B
SE
-0.405
0.254
-0.408
0.276

0.015

0.017

0.016

0.012

0.008

0.012

0.010

-0.005

0.010

0.280
0.123
0.420
0.235

0.277
-0.162
0.400
-0.200

0.282
0.123
0.421
0.236

0.324

-0.372

0.334

**

0.057

0.120

0.107

0.067

0.077

0.127

0.103 0.100

0.104

-0.020

0.104

*

0.080

-0.161

0.149

-0.058

0.106

-0.179

0.142 -0.131

0.142

-0.100

0.142

*

0.107

-0.034

0.200

0.171

0.142

-0.118

0.193 -0.115

0.191

-0.100

0.191

0.067

-0.121

0.121

-0.246

**

0.086

-0.042

0.120 -0.007

0.119

0.253

*

0.119

0.202

-0.428

0.396

-0.684

*

0.281

0.297

*

0.124 -0.109

0.385

0.241

0.384

0.074

0.091

0.134

-0.038

0.095

0.093

0.129 0.109

0.127

0.016

0.127

0.249
Family identity 0.000
Work-family
**
segmentation
-0.642
preference
Family-work
segmentation
0.072
preference

65

Table 2.5 (cont’d)
Model 1
Dependent
variable:

WF ICT Use
B
-0.101

SE
0.269

Model 2
Emotional
Exhaustion
B
SE
*
0.879
-1.813

Model 3
WF Conflict
B
-0.715

SE
0.624

AECE
AECE x Workfamily
*
*
**
0.069
0.133 0.285
0.095
0.152
0.273
segmentation
preference
AECE x Work*
0.140 0.001
0.100
0.316
family tech use
Work-family
conflict
Emotional
Exhaustion
*
**
*
Constant
0.971
2.426 3.561
1.725
2.476
8.355
2
R
0.623
0.409
0.372
Adjusted R2
0.571
0.316
0.273
*
**
Notes: p≤0.05; p≤0.01;
Omitted category for industry: Other
AECE=After-hours electronic communications expectations

66

Model 4
Emotional
Exhaustion
0.362

*

Model 5
Emotional
Exhaustion
B
SE
B
0.158 -1.454
0.840

-

-

0.140

-

-

0.309

0.488

**

-

0.119 0.472
-

2.415
0.449
0.367

0.131

-0.051

0.132

*

0.133

0.114

0.136

**

0.122

0.196

0.129

-

0.332

-

1.252 6.585

Model 6
Turnover
Intentions
SE
B
-0.044
0.849

**

2.348
0.474
0.385

**

0.091

0.342
2.42
0.356
0.242

Figure 2.3: Simple slopes of the after-hours electronic communications expectations and workfamily ICT use interaction on emotional exhaustion
4
Emotional Exhaustion

3.5
3

High
AECE

2.5
2

Low
AECE

1.5
1
0.5
0
Low WF ICT Use

High WF ICT Use

Table 2.6: Simple slope coefficients and statistical significance for Figure 2.3
1
Slope
Slope SE
T
Level of AECE
3.608
0.16
0.26
0.62
2.76
-0.07
0.74
-0.09
1.912
-0.30
0.13
-2.28
1
Notes: Values represent one standard deviation above the mean, at the mean, and one standard
deviation below the mean.
AECE=After-hours electronic communications expectations

Figure 2.4: Simple slopes of the after-hours electronic communications expectations and workfamily segmentation preferences interaction on work-family conflict
3

Work-family Conflict

2.5
2

High
AECE

1.5

Low
AECE

1
0.5
0
WF Integrator

WF Segmenter
67

Table 2.7: Simple slope coefficients and statistical significance for Figure 2.4
1
Slope
Slope SE
T
Level of AECE
3.608
0.35
0.11
3.04
2.76
0.10
0.43
0.24
1.912
-0.14
0.09
-1.54
1
Notes: Values represent one standard deviation above the mean, at the mean, and one standard
deviation below the mean.
AECE=After-hours electronic communications expectations
2.7

Discussion
Model 2 shows that those who prefer to keep work out of the family domain use ICTs

less for work purposes while at home, but that organizational pressure in to stay electronically
connected to the workplace in the form of perceived after-hours communications expectations
can overpower this preference. Previous literature has used organizational pressure as a predictor
of ICT use (Fenner & Renn, 2010), but an alternative interpretation is to treat it as a moderator of
the relationship between ICT use and stress outcomes. The support of this interpretation suggests
that after-hours communications expectations may be functioning as a job demand by pressuring
employees to use ICTs more for work purposes while at home. For segmenters especially, this
can be problematic. The potentially resulting mismatch between work-family segmentation
preferences and actual work-family segmentation practices is the root of several undesirable
outcomes in following models.
One such outcome is shown in Figure 2.4: integrators' role conflict and emotional
exhaustion is largely unrelated to after-hours communication expectations. Those employees one
standard deviation below average on work-to-family segmentation preferences showed no
discernible difference in work-family conflict regardless of the level of after-hours
communications expectations they perceived. Segmenters, on the other hand, experience higher
levels of work-family conflict when after-hours communications expectations are higher, but not
68

when they are lower. The data suggest that work-family segmentation preferences are associated
with lower work-family conflict in low after-hours communications expectation environments,
but the simple slope of this relationship is only marginally significant in a one-tailed test. These
results are consistent with the explanation that a mismatch between after-hours communications
expectations and work-family segmentation preferences is experienced as a job demand in the
form of role conflict, which consequently is related to higher emotional exhaustion.
Unexpectedly, the interaction between work-to-family segmentation preferences and
emotional exhaustion was fully rather than partially mediated by work-family conflict. This may
be due to Model 4 controlling for the interaction between after-hours communications
expectations and work-family ICT use, which was theorized to also capture the role of afterhours communications expectations as a work pressure job demand. Consistent with this
explanation, the interaction between work-family ICT use and after-hours communications
expectations was unrelated to work-family conflict but directly related to emotional exhaustion.
Despite the unexpected full mediation, the pattern of these interactions together supports the
interpretation that after-hours communications expectations function as a job demand in multiple
ways: they create work pressure, and for segmenters, they also create role conflict.

69

Table 2.8: Bias-corrected bootstrap test of indirect effects of interaction between work-family separation preferences and after-hours
electronic communications expectations on emotional exhaustion, through work-family conflict
Bootstrapped 95%
Mediating
Dependent
Observed
Confidence
Independent Variable
Bias
Confidence
1
Variable
Variable
Coefficient
Internal Type
SE
Internal
Normal2
-0.007 0.276
2
N/A
based
Work-family segmentation
Work0.034 0.314 Percentile
preference by after-hours
Emotional
family
0.135
0.072
0.031 0.309 Bias-corrected
electronic communications
exhaustion
conflict
0.012
Bias-corrected
expectations
0.023 0.293 and
accelerated
1
Notes: 5,000 replications were used.
2

Normal-based confidence intervals do not account for bias and are provided for comparison.

Figure 2.5: Empirically supported model
WF
Segmentation
Preference

WF Conflict

AECE

WF Tech Use

Emotional
Exhaustion

Turnover
Intentions
Emotional
Exhaustion

70

Still, the results suggest that after-hours communications expectations, work-family ICT
use, and work-family segmentation preferences may have a more complex relationship than
hypothesized. According to Model 2 and the simple slope analysis shown in Figure 2.3 and
Table 2.6, there was a negative main effect of after-hours communications expectations on
emotional exhaustion, which became more positive as work-family ICT use increased. This
surprising finding may not be of practical concern given that low levels of work-family ICT use
are unlikely in high after-hours communications expectation climates. Regardless, this
relationship is counter to the study predictions because low ICT users in higher after-hours
communications expectations appear to experience lower levels emotional exhaustion. This
finding may be due to controlling for the interaction of after-hours communications expectations
and work-family segmentation preferences in the same model. However, given that this
interaction was also statistically significant, it should not be excluded from analyses and one
interaction should not be interpreted without considering the other. Instead, it may be the case
that these three variables interact together, and accounting for this interaction might facilitate a
clearer interpretation of the pattern of results.
To make sense of this issue, I conducted a post-hoc exploratory analysis, adding the
three-way interaction between work-family ICT use, after-hours communications expectations,
and work-to-family segmentation preferences to the emotional exhaustion model. This model is
shown in Table 2.9. Figure 2.6 shows the simple slopes for the three-way interaction, which is
statistically significant in the regression model (β=0.404; p<0.05). Table 2.10 provides
Bonferroni-corrected tests of the differences between the simple slopes, as recommended by
Dawson and Richter (2006). The Bonferroni correction is a conservative correction for post-hoc
comparisons of simple slopes. Despite the conservative correction, there is a significant

71

difference between the relationship of work-to-family ICT use and emotional exhaustion for
segmenters in high and low after-hours communications expectation climates. For segmenters in
organizations with high after-hours communications expectations, more work-to-family ICT use
is associated with more emotional exhaustion. For segmenters in organizations with low afterhours communications expectations, more work-family ICT use is associated with lower
emotional exhaustion.
This finding suggests that unpressured use of ICTs for work purposes at home may be
beneficial even for segmenters. Though the survey questions were designed to assess technology
based integration of the work role into the family domain, it may be the case that segmenters
used ICTs for work purposes at home to allow themselves to mentally disengage from work, and
that this usage was also captured by the work-family ICT use measure. For example, a segmenter
might find more comfort at home knowing that he/she can be reached in the case of a work
emergency, or knowing that there are no critical e-mails awaiting a response. In this way,
technology can provide justification to mentally disengage from work while at home. In at least
one qualitative study, employees were shown to be able to use ICTs to help segment work and
family roles (Golden & Geisler, 2007). Further research will be needed to better understand how
segmenters and integrators use ICTs differently and whether this is related to individual
characteristics such as personality or family demands.

72

Table 2.9: Post-hoc regression analysis with three-way interaction
Dependent variable:
Emotional Exhaustion
B
SE
1
-0.464
0.258
Manufacturing
1

Service
Work hours at home
Work hours at work
Age
Married
Kids
Elder care
Female
Work-family ICT use
Family-work ICT use
Autonomy
Work identity
Family identity
Work-family segmentation preference
Family-work segmentation preference
AECE2
AECE2 x Work-family separation
preference
Work-family segmentation preference x
Work-family ICT use
Work-family ICT use x AECE2
Work-family segmentation preference x
AECE2 x Work-family ICT use
Constant
R

2

Adjusted R
*

-0.529

0.286

**

0.016

*

0.013

.049
.025

-0.026
-0.156
-0.032
-0.414
0.427
2.892
0.130
-0.140
-0.012
-0.106
3.507
0.129
5.381

*

-1.484
-.95

*

-1.320
.404

*

-8.316

2

*

0.011
0.281
0.128
0.436
0.239
1.569
0.107
0.147
0.197
0.119
1.641
0.133
3.468
0.841
0.378
0.735
0.179
7.163
0.438
0.339

**

Notes: p≤0.05; p≤0.01
1
Omitted Category: Other industries
2
After-hours electronic communications expectations

73

Figure 2.6: Three-way interaction of work-family technology use, work-family separation
preferences, and after-hours electronic communications expectations on emotional exhaustion
5
4.5

Emotional Exhaustion

4

(1) High AECE, High
WF Seg. Pref

3.5

(2) High AECE, Low
WF Seg. Pref

3

(3) Low AECE, High
WF Seg. Pref

2.5

(4) Low AECE, Low WF
Seg. Pref

2
1.5
1
Low WF ICT Use

High WF ICT Use

Note: AECE=After-hours electronic communications expectations
Table 2.10: Bonferroni-corrected simple slope comparisons for three-way interaction of workfamily technology use, work-family separation preferences, and after-hours electronic
communications expectations on emotional exhaustion
Bonferonni
Pair of slopes Compared
p
Corrected p
High AECE/High WF Seg. (1)
High AECE/Low WF Seg. (2)
0.133
0.801
High AECE/High WF Seg. (1)
Low AECE/High WF Seg. (3)
0.002
0.011
High AECE/High WF Seg. (1)
Low AECE/Low WF Seg. (4)
0.220
1.323
High AECE/Low WF Seg. (2)
Low AECE/High WF Seg. (3)
0.754
4.525
High AECE/Low WF Seg. (2)
Low AECE/Low WF Seg. (4)
0.314
1.882
/
Low AECE High WF Seg. (3)
Low AECE/Low WF Seg. (4)
0.291
1.746
Note: AECE=After-hours electronic communications expectations

74

Considering the study results and the exploratory analysis, integrators seem to fare better
than segmenters all-in-all in a knowledge-based work environment. Though segmenters exhibit
similar levels of emotional exhaustion and work-family conflict as integrators when after-hours
communications expectations are lower, they experience more role conflict than integrators when
after-hours communications expectations are higher. This is evidence that the reason for workfamily ICT use is the key to its relationship with employee well-being; segmenters look more
like integrators in their ICT usage patterns when after-hours communications expectations are
high, but only they experience undesirable outcomes. Similarly, segmenters and integrators both
appear to benefit from higher work-family ICT use when after-hours communications
expectations are lower. The worst (i.e. least desirable) outcomes are for segmenters with high
levels of work-family ICT use in climates of higher after-hours communications expectations.
Taking into account the post-hoc analysis, in no case are higher after-hours
communications expectations beneficial to employees' well-being. Organizations can help reduce
stress and potential turnover of employees by allowing them to use ICTs to manage their workfamily roles at their preference rather than pressuring them to use ICTs in a certain way.
Alternatively, employers who value constant connectivity should strive to hire integrators and let
them integrate their roles naturally rather than through organizational pressure. Organizations
that pressure employees to use technology for work purposes in the family domain risk creating
additional job demands of work pressure and role conflict, which are ultimately linked to greater
emotional exhaustion and turnover intentions. Another potential solution for employers that
value constant connectivity is to allow employees to work more from home and reduce
workplace hours. Though work-family segmenters will still suffer when after-hours
communications expectations are higher, they may be able to better manage their conflicting

75

demands with the additional temporal boundary control. For example, work-family segmenters
may be able to chunk their time into work and family blocks more efficiently when they can
work from home.
These recommendations may not apply to employees in jobs that do not primarily require
knowledge work or are more constrained by outside forces beyond the organization’s control.
For example, police officers and doctors may be in constant connectivity to their employers,
especially while on call, because the nature of the job requires someone to always be working. In
these cases, some degree of control is likely necessary for successful completion of job tasks and
organizational success. Despite this, work-family segmenters may still fare more poorly than
integrators on average in terms of stress outcomes in any job which requires more work-family
ICT use. More research is needed to understand whether the effect of expectations are less
detrimental to employee well-being when employees understand that integration is in the nature
of the job and that communication expectations are not a function of the organizational culture.
Similarly, whether segmenters self-select into jobs which do not require such integration remains
an empirical question.
2.8

Limitations and Conclusions
The study suffers from a few limitations. One potential limitation is that the data was

collected from a single source using a survey design. Thus, common method bias may account
for relationships between study variables (Podsakoff et al., 2003). But, there is little evidence
that this was problematic. Several precautions were taken to avoid common method bias,
including a two-wave design and varying scale end points. Additionally, as discussed above, a
Harmon’s Single Factor test failed to find evidence of substantial common method bias. On a
related note, the sample included two waves but cannot be considered longitudinal. More
76

research is needed to understand whether technology-based interruptions are time sensitive as
research suggests that boundary management preferences may vary over time (Kossek &
Lautsch, 2012; Kossek et al., 2012). More research is also needed to examine whether the
direction of ICT use (work-family or family-work) affects employees differently. For example,
work-family ICT use may be less exhausting for those who can compensate with family-work
ICT use as well.
The sample was also relatively small, though this only makes the tests of the hypotheses
more conservative. It does however prevent the use of more sophisticated analytical techniques
such as structural equation modeling. While such a technique would have helped account for
measurement error in the study, the high alphas suggest that measurement error was not
problematic. Additionally, the sample consisted solely of highly educated HR managers. Though
this may be fairly representative of knowledge-based occupations, researchers should be cautious
applying the results to other types of occupations, especially those in which work is less
amenable to being performed via ICTs.
Another limitation is that the three-way interaction of key study variables was not
formulated a priori but rather in response to a surprising finding. Though the slope comparisons
were Bonferroni-corrected, future research should consider this complex relationship carefully.
Finally, a potential problem is that a person’s segmentation preferences may affect their
likelihood to perceive an organization as having high after-hours electronic communications
expectations. The correlation matrix in Table 2.4 suggests though that this is not the case.
Future research should seek to better understand what individual differences impact the
way individuals use and experience technology for work purposes at home. In addition to workfamily technology use, researchers should investigate family-work technology use, i.e. the use of

77

ICTs to meet family demands while in the work domain. Because of its potential impact on
work, family-work ICT use warrants greater research attention and is likely closely related to
work-family ICT use. Future studies should also consider more complex patterns of
segmentation preferences, as some studies suggest that segmentation and integration do not fall
on a continuum and other work-family boundary management styles are possible (Kossek &
Lautsch, 2012). Similarly, future research should examine how after-hours communications
expectations are related to broader employee perceptions of boundary control.
Despite these limitations and unanswered questions, the study supports the idea that
information and communicative technology is a tool that can either help or hinder employee
well-being depending on the circumstances of its use. Generally, there are no benefits associated
with higher after-hours electronic communications expectations, and results were consistent with
their interpretation as a job demand in the form of work pressure and source of role conflict.
Integrators were largely immune to the undesirable effects of these job demands. In order to
prevent emotional exhaustions and associated turnover intentions, employers should encourage
employees to stay technologically connected to their workplace without making it an expectation
or requirement, or they should allow employees greater flexibility in working time and location
so as to lessen conflict with the family role.

78

3

Filling the Childcare Gap: Work-Family Practices and Turnover in Germany

3.1

Introduction
The previous study examines an informal aspect of the way organizations respond to

employee work-family needs. It uses information and communicative technology as a lens
through which to examine the influence of organizational culture on employee work-family role
management and stress outcomes. The study presented in this chapter examines the issue of the
organizational work-family response from a much different perspective. Rather than assessing
employee-level outcomes, it assesses establishment-level outcomes. Rather than examining
informal culture, it examines formal human resource management programs.
Although the second study is quite different in scope, it builds upon the first. The second
study highlights the variety and complexity involved in the rationale behind an organization’s
work-family policy decisions, including elements of public policy environment, employee
demands, and organizational human resource management strategy. It also shows the depth of
consequences for an establishment’s work-family practices; while the previous study measured
individual level turnover intentions, this study measures actual turnover at the establishment
level.
Importantly, the second study also highlights the role of national context in determining
an appropriate work-family policy or practice (Berg, Kossek, Baird, & Block, 2013; Den Dulk,
Groeneveld, Ollier-Malaterre, & Valcour, 2013; Ariane Ollier-Malaterre, Valcour, Den Dulk, &
Kossek, 2013; Piszczek & Berg, forthcoming). Thus, it is appropriate to begin with a discussion
of the contextual issues relevant to this study, which utilizes a large sample of German
establishments. The next section introduces the government-run public childcare system in
Germany and demonstrates how it has created an opportunity for establishments to benefit from

79

a different way of responding to employees’ work-family needs, namely by offering childcare
practices of their own.
3.2

Germany and the Business Case for Childcare
The German government has a federal ministry with an explicitly outlined responsibility

to serve the family role. Historically, family policy in German has been supportive of traditional
gender roles when it comes to work and family, with women abstaining from work in order to
raise children (Abrahamson, 1999; Institute for Child and Family Policy, 2008). A woman who
does not provide full-time care to her children might be criticized as a Rabenmutter, literally
translated as “raven mother,” for metaphorically pushing her chicks out of the nest in order to
work (Bennhold, 2010). Some research suggests that the decision by approximately one third of
women in Germany not to have children is due to difficulty balancing work and family as many
women do want careers (Peus, 2006). Other research has argued that Germany’s past daycare
policies made it difficult for women to participate in the labor market (Hank & Kreyenfeld,
2003; Kreyenfeld & Hank, 2000). Faced with decreasing fertility rates and a potential population
decline, Germany has both reassessed its family policy and begun a push for labor force
participation of women out of concern that the current economy may be unsustainable without
them. As a result, more and more mothers are seeking employment.
With the slowly changing cultural norms and assessment of the labor market, Germany
has simultaneously experienced growing demand for childcare over the past decade. The German
government provides public daycare, but it has not been able to keep up with demand from
parents as childcare grows more socially acceptable, especially for younger children.
Furthermore, even for older children, government provisions are limited to three morning hours
for children aged three to six in most kindergartens. In 2004, 90% of children aged three to six
were enrolled in such programs (OECD, 2004). Children under 3 have been much less likely to
80

be in government-provided childcare services, with only 8.5% participation in 2002 (OECD,
2004). The Rabenmutter stigma applies more harshly to mothers of young children. Even for
older children, however, school typically ends at lunch time.
This is in part due to societal pressure to avoid daycare for young children, but the supply
of early childhood care is still insufficient. As the government has enacted policies seeking to
keep women in the workforce, the stigma of the Rabenmutter has been in decline. Data suggest
that, despite the stigma associated with being a working mother of a young child, many mothers
would use government childcare services if they were available. One study estimated that in
2005 approximately 50% of children under three and 10% of those between ages four and six
were wait-listed for publically subsidized childcare (Wrohlich, 2005), indicating a strong unmet
demand for public childcare among working mothers. This may be the case especially for young
mothers, as the proportion of female college graduate and labor market entrants grows. The lack
of childcare for children under 3 and lack of flexibility in childcare options for older preschool
children have made it difficult for parents—primarily women—to maintain full time jobs.
Because of the incompatibility of full-time employment and existing public childcare
systems, many parents (mothers in particular) find these systems to be inadequate. Most women
report having a hard time finding adequate childcare options (Bundesministerium fuer Familie,
Senioren, Frauen und Jugend, 2007). According to 2010 government data, about two thirds of
women take advantage of out-of-home child care services for at least part of their regular child
care needs, and 53% of young mothers reported that they would take a public day-care spot if
one were available (Bundesministerium fuer Familie, Senioren, Frauen, 2010).
In response to this demand, the German government has acted in two primary ways. One
way is to provide allowances for parents taking care of children. Parents are entitled to
Kindergeld, which consists of flat monthly payments ranging between 184-215 Euros per child

81

depending on how many children are in the household, granted until age 18 (or 25 if the child is
involved in an apprenticeship) (Bundesministerium fuer Familie, Senioren, Frauen, 2013b). In
2007, Germany began the Elterngeld parental work leave program, which provides up to 1,800
Euros per month up to 14 months divided between both parents (or fully for an unmarried parent)
with a maximum of 12 months per parent (Bundesministerium fuer Familie, Senioren, Frauen,
2013a). In 2007 and 2008, 87% of mothers claimed the maximum 12 months (Boettcher &
Krieger, 2008). In 2013, new legislation resulted in another monthly subsidy called
Betreuungsgeld to care for one’s own young child aged one or two at home and abstain from
using government daycare services (Bundesministerium fuer Familie, Senioren, Frauen, 2013c).
These practices may help ease the burden on the public system, but critics argue that
Betreuungsgeld dissuades labor market participation of women. They argue that it instead
encourages a male breadwinner model in which women stay home to raise children, taking the
subsidy rather than enrolling their young children in public daycare and continuing to work
(Hawley, 2012; Heidenreich, 2013).
The second strategy utilized by the government and arguably more conducive to the labor
force retention of mothers was to expand public daycare systems. The government passed the
Childcare Expansion Act in 2005 which created more public daycare spots, tripling them over
the course of several years. But in 2010, the demand for early childhood education was still
growing and remained unmet (The Local, 2010). More recently, public daycare has been opened
to children as young as one year old (as opposed to three), but in 2013, the government estimates
there is still a shortage of 220,000 childcare spots for young children (Pauls, 2013). This shortage
is in part responsible for the introduction of Betreungsgeld, which is expected to reduce demand
for daycare spots but at the potential risk of female labor force participation.

82

The subject of publically provided and subsidized childcare continues to be a
controversial issue in Germany as values adapt to labor market demographic shifts resulting from
low fertility rates and increasing proportions of female college graduates looking for work.
While the government continues to seek solutions, there is another viable option for addressing
childcare needs. The high demand for childcare and lack of sufficient public childcare options in
Germany offers German establishments an opportunity to fill this gap by offering practices such
as onsite childcare or subsidies for private childcare. These practices, whether in the workplace
of the mother or the father, allow for both parents to remain active in the work role after the birth
of a child. This may also help with low fertility rates considering that, as mentioned above,
research suggests many women that choose not to have children do so because they want to be
more successful at work. These types of childcare in which the parents retain some physical
proximity to their children during working hours may also be more socially acceptable to the
general public, potentially diminishing the Rabenmutter stigma. They are also easier on
employees than public daycare options because they reduce commuting time, are more likely to
be flexible around working hours, and may allow parents to easily see their children during
working time if necessary.
Some companies have already taken note of the potential benefits of onsite childcare
facilities; major corporations Bayer, BASF, and Volkswagen all have childcare centers of
varying levels of coverage for employees (BASF, 2013; Bayer, 2013; Stolz, 2010). Siemens
similarly has day care centers near its production sites noting that women make up 21% of its
staff but 34% of its new recruits (Bennhold, 2010). At least one private consulting organization,
Department of Tomorrow, is attempting to actively promote corporate social responsibility in
other German companies by helping them establish onsite daycare centers. As stated by one of
its consultants, Dominik Schmengler, “Employees have to work a lot, but there is the family they

83

want to take care of and there’s a demand by the society to find a solution. How can we fit both
demands together? And the political answers are not enough at the moment” (Pauls, 2013).
Schmengler also mentions that practices such as these are a “pull factor” in attracting good
employees.
This conventional wisdom is in fact consistent with academic theory. In addition to
helping employees with families from a humanistic perspective, strategic human resource
management (HRM) theory suggests that establishments can also benefit from employee-friendly
work-family practices concerning childcare. The strategic HRM literature argues that HRM
practices can help organizations generate a sustainable strategic advantage over their competitors
(Delery & Doty, 1996; Huselid, Jackson, & Schuler, 1997; Huselid, 1995; Offstein, Gnyawali, &
Cobb, 2005). Such practices typically include compensation systems, training programs, or
performance review programs which are more directly linked to employee performance (Becker
& Huselid, 1998; Combs, Liu, Hall, & Ketchen, 2006; Evans, 2005; Gittell, Seidner, &
Wimbush, 2010). When employee performance is increased in aggregate, the organization reaps
the benefits of increased effectiveness or productivity. Research, discussed below, suggests that
work-family practices (WFPs) like onsite or subsidized childcare may also help organizational
performance, albeit less directly.
WFPs are organizational practices that allow employees to better manage potentially
competing work and family demands (Kossek & Michel, 2011) and are thought to have strategic
value through increasing employee attraction to and retention in organizations (Carless &
Wintle, 2007; Honeycutt & Rosen, 1997; Kossek & Nichol, 1992; B. Lee & DeVoe, 2012; Rau
& Hyland, 2002; Richman et al., 2008). WFPs come in many different forms, though in thus
study I focus on childcare-related practices due to their importance in the German context. Other
types of practices may not function the same way in this context, whereas childcare-related

84

practices similarly may not function the same way in other contexts where public childcare is
less in demand.
Though strategic HRM studies show a consistent link between human resource practices
and organizational performance, including WFPs, these studies suffer from an inability to make
causal inferences. This issue remains largely unresolved in the literature and poses a threat to its
fundamental argument that HRM practices can increase organizational performance. As will be
discussed below, some researchers plausibly argue that better-performing organizations may be
better positioned to offer more employee-friendly HRM practices (Huselid & Becker, 1996;
Wright, Gardner, Moynihan, & Allen, 2005). This issue remains largely unresolved in the
broader strategic HRM literature, and research that accounts for this issue is even rarer in the
domain of work-family HRM practices. Thus the goals of the present study are to first examine
whether the offering of childcare-focused WFPs is associated with lower levels of turnover in
German establishments, and second to do so using an analysis that provides a clearer
understanding of the nature of this relationship.
Regarding the first goal, examining the strategic value of WFPs is becoming increasingly
important. As organizations compete more globally and labor market demographics shift, the
strategic HRM literature must generate richer contextual analyses in order to bring research in
line with current business needs and expand theory to accommodate environmental factors that
may affect HRM strategy formulation and effectiveness (Kaufman, 2010). Similarly, workfamily researchers have called for richer contextualization and theorization at the national level
(Ariane Ollier-Malaterre et al., 2013). Regulative institutions such as public policy can impact
whether and to what extent certain HRM practices are offered as well as the extent of their
strategic value (Berg et al., 2013; Piszczek & Berg, forthcoming). Though the strategic HRM
literature and the broader work-family literature include many international studies, many of

85

them do not fully explore the contextual influences of the regulative and cultural environment on
the strategic effectiveness of certain HRM practices. I discuss how legal benefits and culturallydriven demand surrounding childcare affect the potential strategic effectiveness of HRM
practices and the generalizability of the results of the present study to establishments in other
countries.
Regarding the second goal, the present study’s unique analysis provides more insight into
the nature of the relationship between strategic HRM practices and organizational outcomes. As I
explain in greater detail below, it is unclear whether organizations that adopt strategic HRM
practices experience a subsequent increase in organizational performance or better-performing
organizations are more likely to adopt strategic HRM practices. I examine the relationship
between the presence of childcare practices both between establishments, and also within single
establishments as their own childcare practice offerings change over time. Though this analytic
technique does not solve it, it provides some insight into the causality controversy and eliminates
several potential theoretical explanations for relationships observed in strategic HRM research.
This study also provides several empirical contributions, above and beyond being one of
the few to examine WFPs from a strategic HRM perspective and address the issue of causal
direction. First, the study provides a more generalizable analysis of the relationship between
establishment childcare practices and turnover. Studies with large, multi-industry samples are
needed to provide a more balanced perspective on the prevalence and strategic effectiveness of
WFPs, particularly childcare which has received relatively little research attention compared to
more common WFPs like flextime. I use data from a nationally representative annual panel
dataset of approximately 16,000 German establishments collected by the German government
Institute for Labor Market and Occupational Research (IAB) which includes information about
childcare offerings. Second, the study splits the variation in the dependent variable—turnover—

86

into within and between establishment components. This allows for a clearer empirical
understanding of the source of turnover’s variation and separates the issue of which
establishments offer childcare practices from childcare practices’ relationship to employee
turnover. Third, I am able to examine employee turnover by gender. Given the cultural pressure
for women to take time off work to care for their children in their first years and the increasing
female labor force participation rates in Germany, WFPs may work differently for men and
women. Finally, most studies examining the strategic value of WFPs do not measure actual
turnover at the organization level but rather turnover intentions at the employee level. I use an
establishment-level measure of actual turnover as the key dependent variable. The next section
provides a discussion of the literature studying the role of WFPs in strategic HRM.
3.3

Work-Family Practices and Strategic Human Resource Management
The modern strategic HRM literature is heavily influenced by Huselid (1995). The

relationship between HRM practices and firm performance is the foundation for the strategic
HRM literature and Huselid’s study is the most widely cited (Fernandez-Alles & RamosRodríguez, 2009). Using a large sample of firms, Huselid demonstrated that HRM practices were
negatively related to turnover and positively related to productivity and corporate financial
performance. Following this study, strategic HRM practices have been widely shown to have
positive associations with various measures of organizational performance. One meta-analysis of
92 studies demonstrated a clear positive relationship between the adoption of high-performance
work practices—a subset of strategic HRM practices—and organizational performance (Combs
et al., 2006).
Work-family practices (WFPs) are also a subset of HRM practices, though they are not as
commonly considered from a strategic HRM perspective. These are practices (sometimes called
flexible work initiatives or work-family provisions) which help employees manage their work
87

and family role demands (Kossek & Michel, 2011). Their central feature is that they give
employees some degree of control over when, where, or how they work with the most proximal
goal being the improvement of employee work-family role management. WFPs can thus take
several forms which alter the timing, location, amount, or distribution of work (Kossek &
Michel, 2011). Various such WFPs concern childcare; two of these are the focus of the present
study. First, onsite and subsidized childcare allow employees to focus on work demands during
time designated for the work role, minimizing family interruptions and distractions (Payne,
Cook, & Diaz, 2012). Childcare also allows employees to work who might not be able to due to
a lack of other childcare options. Second, allowing employees to stay in touch with the
workplace during designated parental leave (henceforth called parental leave contact) slows the
depreciation of human capital associated with taking substantial time off from work. This policy
should allow an easier transition back to the workplace after parental leave has ended.
WFPs such as these may help increase employee performance by minimizing familybased interruptions of work for those with high family demands (Shockley & Allen, 2007) or
keeping employees on leave up-to-date with establishment issues and practices. But while other
HRM practices may improve organizational performance through aggregate improvements in
individual employee performance (Combs et al., 2006), the primary strategic value in WFPs is
thought to come from their ability to attract talented job applicants or retain existing talented
employees, particularly from demographic groups which tend to place a higher value on
flexibility (Berg, Kalleberg, & Appelbaum, 2003; Glass & Estes, 1997; Osterman, 1995). A
review by Beauregard and Henry (2009) breaks down the literature on the effects of WFPs on
organizational strategic interests and summarizes the results of both individual and
organizational outcomes of studies of the adoption WFPs. They conclude that WFPs are effective

88

in the recruitment and retention of employees because WFPs signal broader organizational
support for employees, fostering commitment (e.g., Casper & Buffardi, 2004).
On the other hand, the review finds mixed evidence that WFPs are associated with
increased firm performance more directly, like other strategic HRM practices. Perry-Smith and
Blum (2000) for example found that more WFPs were associated with higher perceptions of firm
performance. But, Bloom, Kretschmer, and Van Reenen (2011) found that any benefit of WFPs
on firm productivity can be explained by a more general “quality of management practices”
measurement. Similarly, studies by Konrad and Mangel (2000) and Lee and DeVoe (2012) found
that WFPs are beneficial to some organizations but not others. Evidence for a substantial impact
of WFPs on firm-level productivity is thus at this point inconclusive.
Beauregard and Henry suggest that investigating other mechanisms through which WFPs
can impact firm performance, such as employee turnover levels, may provide a better way of
examining the effect of WFPs on firm performance than examining their effects on employee
productivity. In support of this contention, many individual-level studies grounded in
organizational behavior perspectives have shown that the availability and use of WFPs are
associated with greater employee commitment and other desirable outcomes. One recent metaanalysis concluded that greater WFP availability was positively associated with higher affective
commitment, higher job satisfaction, and lower work-family conflict (Butts et al., 2013). Other
studies have shown that an employee’s affective commitment is linked with lower turnover
intentions and absenteeism (Somers, 1995). Though these studies provide empirical support for
pieces of the chain between WFP availability, organizational commitment, and desirable
organizational outcomes, some studies have examined the link between WFP availability and
organizational attraction or employee turnover intentions directly.

89

These studies have shown that job applicants are more attracted to firms offering more
work-family practices (e.g., Carless & Wintle, 2007; Honeycutt & Rosen, 1997; Rau & Hyland,
2002). Similarly, many studies have shown a negative link between WFP availability and
turnover intentions and absenteeism (Baltes, Briggs, Huff, Wright, & Neuman, 1999; Dalton &
Mesch, 1990; Ralston & Flanagan, 1985). The above-mentioned meta-analysis showed an
overall positive relationship between WFP availability and intentions to stay across 19 studies
(Butts et al., 2013). Batt and Valcour (2003) showed that some work-family practices were
associated with reduced turnover intentions even controlling for other HRM practices such as
benefits and job design. While these studies do not examine actual turnover, provide nationally
representative cross-industry samples, or carefully address issues of causal direction, they
provide empirical support for the effectiveness of WFPs. Taken together, the research on workfamily practices in the strategic HRM domain suggests that one way organizations are likely to
benefit from their presence is through increased retention of employees.
As discussed in depth above, the state of public childcare in Germany provides some
insight into what types of WFPs might be particularly effective in increasing employee retention.
In Germany, because public childcare is in high demand and relatively short supply,
organizational childcare practices may be especially effective in employee retention. Similarly,
social pressure to care for children during their first few years may compel some employees to
take longer parental leaves than desired. Parental leave contact makes returning to work easier
and keeps employees’ skills more relevant during the leave period. This line of research leads to
Hypotheses 1a and 1b. Note that these hypotheses compare different establishments, constituting
between-establishment predictions.
Hypothesis 1a: Establishments with a higher prevalence of childcare practices over time will
have lower overall turnover relative to other establishments on average.

90

Hypothesis 1b: Establishments with a higher prevalence of parental leave contact over time will
have lower overall turnover relative to other establishments on average.
Despite the supportive evidence from the literature outlined above, Beauregard and
Henry also identify several shortcomings in linking WFPs to organizational performance. One
key criticism is that most research consists of case studies, small samples of firms, or individuallevel outcomes. These results cannot necessarily be reliably extrapolated to other firms. For
example, a study by Berg et al. (2003) examined whether firm practices were related to
employee work-life balance perceptions, but only in the manufacturing industry. The abovementioned study by Bloom and colleagues (2011) also only examined manufacturing firms.
Another limitation is a lack of measurement of actual turnover. Most studies have examined only
turnover intentions of individual employees due to the difficulty of studying actual turnover, but
not everyone who intends to quit does so. Evidence suggests the correlation between intentions
to quit and actual quitting is about 0.5 (Mayer & Schoorman, 1992; Steel & Ovalle, 1984). This
prevents a true measurement of organization-level turnover critical to the strategic HRM
perspective. In addition, turnover intentions are commonly measured at the individual level and
thus subject to problems associated with aggregation and sampling within an establishment.
They are consequently rarely extrapolated to the organizational level, even though strategic
HRM researchers are interested primarily in organization-level outcomes. As a result, most
research examining the relationship between WFPs and turnover is focused on individual
turnover intentions rather than actual organizational turnover. Importantly, this research also
does not address issues of causal direction between policy adoption and turnover.
Another shortcoming of the current literature is the lack of research on many specific
WFPs. Given the large number and many types of WFPs and their collective unclear relationship
with organizational performance, little research has provided a nuanced perspective on many

91

specific WFPs. In many cases, WFPs are indexed in analysis, but all WFPs are not necessarily
equal and the effectiveness of a particular WFP may depend on contextual factors such as the
public policy environment. Though more common practices like flextime and parental leave are
well-represented in the literature, only a handful of studies have examined the strategic benefits
of childcare practices. Those that have done so have had limited samples and measured
intentions to turnover rather than actual turnover. Batt and Valcour (2003) found no relationship
between dependent care policies and turnover intentions, but their sample included only whitecollar employees in dual-career families who may be better able to afford childcare on their own.
Similarly to the broader collection of WFPs, childcare practices specifically are thought to be
more effective at increasing attraction and retention of employees than having more direct effects
on employee performance (Kossek & Nichol, 1992). Payne, Cook, and Diaz (2012) found that
access to quality childcare was associated with lower turnover intentions and absenteeism
through reduced family interference with work.
Complicating matters further, the relationship between WFPs and turnover may vary
from society to society (Stavrou & Kilaniotis, 2009) suggesting that a much more careful
perspective is needed to understand why and how individual WFPs are strategically viable. This
also supports the need for having studies of specific WFPs, as certain WFPs may be better suited
to particular contexts. In Germany, where childcare demands are particularly high and public
options are not readily available, organization-sponsored childcare practices may be especially
attractive to employees and thus more strategically viable for organizations in terms of their
ability to attract and retain employees (Piszczek & Berg, forthcoming). This may also explain the
lack of research surrounding parental leave contact policies, which may be more commonly
informal in the United States where the majority of WFP studies draw their samples. It may
however be more important and thus more formalized in contexts such as Germany with stronger

92

sociocultural pressure for mothers to stay at home with children for several years and longer,
paid maternity leaves.
I address these issues in the present study. I use a large, nationally representative dataset
which covers all industry classifications in Germany with data on actual turnover at the
establishment level. I examine childcare and parental leave contact practices specifically and
separately, and above argued why in the context of Germany they are particularly likely to be
successful. Later, I make comparisons to other cultural and institutional contexts. One final issue
I address is a bigger problem in the broader strategic HRM literature alluded to above: the nature
of the relationship between strategic HRM practices and organizational performance.
3.4

Conflicting Effects in Strategic Human Resource Management Theory
In order to develop the next pair of hypotheses, the relationship between HRM practice

and organizational performance must be carefully considered. Strategic HRM theory posits that
the adoption of a strategic practice will subsequently lead to improved organizational outcomes,
but some scholars suggest the reverse is also plausible: organizations with better performance are
better positioned to adopt more and better quality HRM practices (Wright et al., 2005). This may
be the result of either having more financial support for such practices or as a reward to
employees for their performance. Though the strategic HRM literature makes a strong theoretical
case for the practice-to-performance causal direction, there is empirical evidence to suggest that
the performance-to-practice effect is in play, or even that both effects are operating
simultaneously.
One detailed description of this dilemma comes from a study by Wright, Gardner,
Moynihan, and Allen (2005). The study critiques the strategic HR literature for a lack of
“rigorous designs to test the hypothesis that employing progressive HRM systems actually
results in higher organizational performance in a causal sense.” While some studies did test their
93

hypotheses longitudinally, many did so using retrospective designs in which participants needed
to remember the practices of their organization in the past (e.g., Guthrie, 2001; Ichniowski &
Shaw, 1999). Through partial correlations, the Wright et al. study demonstrates that the adoption
of HR practices was related not only to future performance, but also past performance.
Another methodologically rigorous study provides evidence that practice and
performance may have a reciprocal relationship. Van Iddekinge et al. (2009) used a latent growth
modeling approach to show that, over time, the relationship between strategic HRM practice and
organizational performance was reciprocal. They also showed that different policies and
practices operated in different causal directions. For example, in their sample they found that
training practices drove organizational performance, but organizational performance drove the
use of selection procedures. Because of this, examining relationships between specific strategic
HRM practices and organizational performance individually may be even more important.
Finally, in their meta-analysis of high-performance work practices, Combs, Lio, Hall, and
Ketchen (2006) suggest that practices and performance operate in a feedback loop which authors
should consider in future research.
Thus, causality remains a key area for further research in the strategic HRM literature. In
order to make a strong argument for causality, research must meet three criteria (T. D. Cook &
Campbell, 1979): covariation between cause and effect, temporal precedence of the cause, and
elimination of alternative explanations for the possible cause-effect connection. By
demonstrating a statistically significant relationship between two variables, as the majority of
studies have in this literature, only the first criterion is met. The second can only be addressed
with longitudinal data collection which measures the dependent variables at different times as the
independent variables fluctuate. In order to meet the third, an expansive dataset is needed to rule
out potential confounding effects, which at the firm level are numerous yet difficult to measure.
94

Researchers typically lack the resources to be able to survey large numbers of organizations at
all, let alone longitudinally or with large surveys which cover many potential controls. Because
of these issues, causality has been difficult to establish in macro-level studies focused on
organizations as the subject of interest.
One way to potentially better understand the nature of the relationship between strategic
HRM practices and organizational performance is to examine it both between establishments
with different HRM practices and within establishments as they change their HRM practices
over time. Though this technique does not meet all the requirements for causality outlined above,
as both components are still cross-sectional in nature, it does provide more insight into the
relationship in question. While a between-establishment analysis compares the turnover levels of
different establishments based on whether they had a WFP in place, the within-establishment
analysis uses longitudinal data to examine the average relationship between turnover and the
presence of a WFP in a single establishment, as that practice changes over time, holding constant
any time-invariant aspects of that organization. With this analysis, researchers can answer two
different questions: are establishments with strategic HRM practices better performers, and is a
change in HRM practices associated with a change in organizational performance?
The first question, represented by Hypotheses 1a and 1b above, is addressed by more
typical, cross-sectional between-establishment relationships. Because one cannot infer causation
from correlational relationships such as cross-sectional regression coefficients, these
relationships are consistent with multiple theoretical explanations. For example, fundamental
strategic HRM theory suggests that a negative relationship between presence of a WFP and
turnover in an establishment is because the adoption of the practice caused a decrease in
turnover. Alternatively, an establishment with very low turnover may adopt WFPs to reward its
committed employees or because they have lower employee replacement costs. Because cross-

95

sectional studies are typical due to difficulties in collecting longitudinal data, most studies cannot
differentiate between these two causal directions.
To provide more insight into the exact nature of the between-establishment effects, when
data provides, researchers can also examine the relationship between WFPs and performance
within a single establishment as it changes its WFPs over time. This can be done with an
establishment fixed-effects model, which controls for time-invariant characteristics of the
establishment and thus comes much closer to meeting the third criterion for establishing
causality. If there is no relationship between WFP adoption and turnover within establishments,
it may be that they are not effective at retaining employees already in the establishment. If a
negative relationship between WFP adoption and turnover is observed, this provides much
stronger causal evidence as all other time-invariant organization-level explanations for the
turnover change can be ruled out while other potentially time-variant explanations can be
controlled. In this sense, the fixed-effects model is a natural quasi-experiment in which changes
in turnover can be observed alongside changes in work-family practices.
Unfortunately, examining within-establishment effects is a difficult task as
establishments do not change their practices very frequently. Finding a sufficient number of
within-establishment changes in HRM practices using a random sample is even more
challenging. The present study is able to meet these challenges because its sample is comprised
of approximately 16,000 German establishments over several years collected using a stratified
random sampling technique. Even though relatively few establishments changed their WFPs
across the years of the study, as described in the results below, this sample still provides several
hundred instances of such changes, creating a natural quasi-experiment of sufficient size from
which to examine within-establishment effects.

96

Recall that Hypothesis 1 stated that establishments with more prevalent childcare
practices would tend to have lower levels of turnover. This between-establishment effect is
consistent with strategic HRM theory’s prediction that the adoption of a childcare practice will
reduce turnover. Hypothesis 2 is also consistent with this prediction, but is stated in terms of the
within-establishment effect.
Hypothesis 2a: In years when an establishment has a childcare practice is in place, it will have
lower turnover relative to years when it does not have a childcare practice in place.
Hypothesis 2b: In years when an establishment has parental leave contact in place, it will have
lower turnover relative to years when it does not have parental leave contact in place.
Hypotheses 1 and 2 are both consistent with strategic HRM theory but together make a
more powerful prediction. Support of Hypothesis 1 suggests two possibilities. First, as predicted
by strategic HRM theory, childcare and/or parental leave contact practices are effective in
reducing turnover, such that they drop the establishments that have them to lower levels of
turnover than those that do not. This would result in lower turnover among establishments with
these practices being observed in the between-establishment analysis. Second, counter to
strategic HRM theory, establishments with these practices may offer them because they have
lower turnover, either as a reward to dedicated employees or because they are better positioned
financially to create such practices and programs. Because Hypothesis 1 does not differentiate
between the potential directions of the effect, it is a better indication of what types of
establishments have such practices in place rather than what the effect of those practices are.
This is discussed in depth later.
Hypothesis 2 suggests that it is the first of these two potential effects at play, because if
childcare practices were being used to reward employees, turnover would likely be lower
regardless of the presence of a childcare practice and we would not observe lower turnover

97

within a single establishment when it has the practice present and when it does not. In other
words, there would be no within-establishment effect. All together, the present design cannot
provide a full causal analysis of the relationship between HRM practice adoption and
organizational performance, but the inclusion of a within-establishment component provides
evidence to eliminate some conflicting explanations and make a stronger argument about the
direction of the relationship. Aside from causal issues, however, another important consideration
is whether the effects of these practices may disproportionately affect women.
3.5

WFP Effectiveness and Gender
As mentioned above, Germany’s federal government recently passed a law providing a

monthly childcare allowance for children between ages one and three called Betreuungsgeld.
This policy essentially extends the widely used one year maternity leave until a child is age
three, at a government cost of about 1.2 billion Euros annually (Pauls, 2013). Critics contend this
small subsidy will encourage women to become stay-at-home moms rather than pursue a career,
and that the money would be better spent further expanding the public daycare system.
Proponents say that these mothers were already likely to have stayed at home rather than
returning to the labor market. Though it is too early to tell the effects of this policy on female
labor market participation, its effect on the employment of women is a key issue of interest. This
may also be of importance to establishments. If critics are correct, the new policy should make
establishments with childcare practices even more attractive to women who wish to stay in the
labor market. Childcare practices allow employers to compete with the Betreuungsgeld
allowance. For mothers for whom the allowance was just enough to keep them out of the labor
market, having a comparable option from a workplace may be sufficient to keep them there.
There is substantial theoretical and empirical evidence from strategic HRM theory that
certain practices may be more effective for particular subgroups of employees. In societies where
98

caregiving demands are a primarily female role, practices that assuage those demands should
have more influence on the employment decisions of women. Indeed, childcare has been
traditionally seen as a female role in Germany (Treas & Drobnic, 2010). Butts and colleagues
(2013) found that WFPs were more strongly related to employee attitudes in samples consisting
of more women and more individuals with family demands. Casper and Harris (2008) found that
the relationship between WFP availability and organizational commitment was unconditional for
women while for men it was dependent on WFP use. Konrad and Mangel (2000) also found
stronger relationships between WFP availability and organizational performance when the
percentage of female employees was higher. Though childcare provisions are likely beneficial
for both men and women, this evidence suggests that they may be more beneficial for women
than men on average. Thus, I expect that childcare and parental leave contact practice prevalence
will be more strongly associated with female turnover than male turnover across both the
between and within-establishment effects described in Hypotheses 1 and 2.
Hypothesis 3a: The relationship between childcare practice offering and turnover will be stronger
for turnover of women than for turnover of men.
Hypothesis 3b: The relationship between parental leave contact offering and turnover will be
stronger for turnover of women than for turnover of men.
3.6

Methods

3.6.1 Sample
Data for the current study comes from the IAB Establishment Panel. The IAB
Establishment Panel is a nationally representative random survey of German establishments
conducted annually on June 30th to based on a random stratified sample. The sampling frame
includes any establishment which employs at least one person subject to social security. The
population of the panel consists of roughly 16,000 establishments each year. Though the surveys
99

typically cover general financial and employment information, each year a supplemental survey
is attached, providing additional questions regarding a specific topic of interest, including the
existence of several work-family practices in some years. Because of this aspect of the survey
design, information regarding WFPs is not available for every year. The present study used data
from 2002, 2004, and 2008, the years which establishment surveys contained questions about
WFPs. Over the three survey years there is some attrition as establishments close or discontinue
participation. Thus, the number of observations varies with each type of analysis and can be
found in the relevant tables presented for each analysis.
3.6.2 Variables
All variables are measured at the establishment level. The first key independent variable
is the presence of an establishment practice which assists employees with childcare. This
variable was assigned a “1” if establishments had any of the following: “Workplace childcare
facilities,” “Sponsoring childcare, support by parents’ initiative,” or “Further assistance
concerning childcare.” In 2002, these three offerings were assessed individually but in
subsequent years they were combined into a single item. I thus recoded the 2002 measure to be
consistent with the other years. Establishments were also asked whether they had a policy of
keeping in contact with employees while they were out on parental leave which was similarly
scored with “1” or “0.” The key dependent variable is establishment turnover. This variable is
the number of employees exiting the establishment for any reason in the first half of the year (as
the survey is conducted on June 30th). The survey reports total turnover and turnover of women;
turnover of males was calculated by subtracting female turnover from total turnover.
In each model, I include several control variables which may affect turnover rates.
“State” is the German federal state where the establishment was located and “industry” is a onedigit industry indicator. Other control variables assess the establishment’s labor force
100

characteristics, including the total number of employees at the establishment, the proportion of
the total number of employees which were female, employees’ average weekly working hours,
and employees’ monthly average gross pay. Establishment profit is also included as a control,
measured using a single Likert scale question reading “Please give your assessment of the profit
situation of your business in the last fiscal year.” Business volume is also included, providing a
control for the financial size of the business. The survey also assesses whether the establishment
offers overtime, operates under a collective agreement, or has a works council. These are all
dichotomously coded as either “1” for the presence or “0” for the absence of each characteristic.
To control for a more general treatment of women in the establishment, dichotomous controls for
a policy of promoting women and absence of any equal opportunity policies are included.
Finally, year controls are included to contain any general trends in turnover that occurred
through the years in the sample.
3.6.3 Analysis
The dataset for this analysis is a large pooled cross-section with observations at the
establishment-year level; each establishment appears as a separate case for each year it appears
in the panel, between one and three times. Results are presented using linear regression models.
All models use standard errors clustered on establishment to account for the non-independence
of observations, which may otherwise bias standard errors leading to spurious results. The
between-establishment model uses a pooled cross-section of all the data (controlling for year as
well as all other above-mentioned controls). These models address Hypotheses 1a and 1b. They
examine the average relationship between having a childcare or parental leave contact policy and
turnover. The within-establishment model specifies establishment-fixed effects, controlling for
time-invariant characteristics of establishments. Time-fixed effects are also included as a control.
Because an establishment’s industry and state rarely vary over time, these controls are
101

encompassed in the establishment fixed effects and thus not modeled directly. This model
addresses Hypotheses 2a and 2b.
I also present a third model which combines the between- and within-establishment
effects into a single estimated model. This provides a robustness check for the results of the
separate models in addition to providing more efficient estimations that, unlike a purely random
effects model, do not use a matrix-weighted average of the between- and within-establishment
results (Gould, 2001). Instead, I calculate the means of each independent variable by
establishment and include them in a random effects model. This model also includes standard
errors clustered on establishment to account for non-independence of the data resulting from
multiple observations for each establishment.
Importantly, the dependent variables are count variables with zero values for
establishments reporting no turnover. Though log transformations are commonly used with such
variables, this usually requires altering observations of zero (e.g., by adding one to each zero
value so the log can be taken from zero values), and such transformations do not address nonnormality of error structure (O’Hara & Kotze, 2010). Although the use of clustered or robust
standard errors can account for the bias due to non-normally distributed dependent variables, the
structure of the turnover variable could potentially be problematic if linear models are used.
Thus, as an additional robustness check, I conduct each of the three analyses using
poisson regression as well. These models, presented in Appendix B, are meant to mirror the
within-establishment, between-establishment, and combined models presented in this chapter.
Poisson models also allow for the use of an exposure variable which accounts for the opportunity
for a particular value of the dependent variable to occur. Because turnover count is likely closely
related to the total number of employees in an establishment in that it creates an upper ceiling,
this was used as the exposure in the poisson models. For the poisson models, incidence rate

102

ratios (IRRs) are also reported, which can be interpreted as a proportional and constant hazard
rate. In other words, the IRR is the proportion change in the turnover count associated with a
one-point change in the independent variable.
Finally, to address Hypotheses 3a and 3b, the coefficients for male and female turnover
are directly compared by calculating a z-score for their difference in magnitude. This comparison
is shown for both the distinct between- and within-establishment models as well as for the
combined model. Additionally, it is shown for both childcare and parental leave contact.
3.7

Results

3.7.1 Variance Components Analysis
An important precursor to estimating both between- and within-establishment effects is to
demonstrate that the dependent variable varies sufficiently at each level. If there is little variance
in a variable at a certain level of analysis, there is also little practical value in assessing its
relationship with predictors at that level. In order to estimate the proportion of variance in the
dependent variables due to within- and between establishment variation, I conducted a one-way
ANOVA using the loneway command in Stata 12. The results are presented in Table 3.1 below.
Each ANOVA gives an ICC statistic, interpretable as the upper bound of the proportion of
variation in each dependent variable explainable by the establishment ID variable. In other
words, it is the upper bound of the proportion of variance explainable by between-establishment
effects. Whatever variance is not captured by between-establishment effects is due to withinestablishment effects and error, which are indistinguishable.
Table 3.1 reveals that a substantial portion of between-establishment variance may exist
in each of the three dependent variables, as represented by the intraclass correlation coefficient
(ICC). This is supported in each model by a statistically significant F-test, testing against the null
that the between-establishment sum of squares is equal to zero. However, the ICCs are not so
103

large as to indicate no additional sources of variation in turnover measures. While approximately
60-65% of variation in each turnover variable can be attributed to between-establishment
differences, the remaining proportion is due to variation within an establishment over time, plus
error. Because this is a government-collected dataset collected alongside information relevant to
taxes and social security, there is not likely to be much measurement or sampling error as
misreporting could have financial implications for the establishment and the sample is large.
Thus there is substantial variation in turnover at both between- and within-establishment levels
of analysis, supporting the discrete modeling of each component and the differentiation of their
effects in Hypotheses 1 and 2.
Table 3.1: One-way ANOVA results on dependent variables demonstrating within-and-between
establishment variance components
Variable
ICC
F
Prob>F
Total Turnover 0.64984 3.68
0.000
Total Turnover 0.64173 3.58
0.000
(Male)
Total Turnover 0.59864 3.15
0.000
(Female)
3.7.2 Descriptive Statistics
Table 3.2 shows the descriptive statistics of the study variables, and breaks down the
standard deviations of each measure into between- and within-establishment components.
Within-establishment versions of each variable are centered at the establishment mean and
therefore, by definition, will have negative minimums and means of zero. Thus only overall
means are shown. Standard deviations for both levels of variation are provided to better indicate
how much variation lies at each level of analysis for each variable. The descriptive statistics
reveal that the average establishment was moderately large with a size of 170.51 employees, but
with an even larger standard deviation (859.5235). On average, establishments reported that 15
employees turned over in the 6 months preceding the survey, the majority of whom were male.
104

Only about 4% of establishments offered childcare practices to employees while 15% offered
parental leave contact (across all years). The within-establishment standard deviations of 0.1 for
childcare and 0.177 for parental leave contact suggest that establishments are largely, but not
completely, consistent in their WFPs, and that parental leave contact practices are a bit more
volatile than childcare practices. Table 3.3 shows the prevalence of these practice changes across
the establishments that participated in all three waves of the panel and confirms that the majority
of establishments never had these practices at all. Table 3.4 shows the distribution of industries
and WFPs among them. Childcare prevalence ranges from 1.39% of observations in construction
to 10.89% in financial intermediation, while parental leave contact ranges from about 5% in
agriculture to 40% in financial intermediation.
Table 3.2: Variable descriptive statistics with between and within standard deviations
Variable
Level
Mean
Std. Dev.
Min
overall
170.5101
859.5235
1
Number of employees
between
816.728
1
within
63.05975
-4057.82
overall
67.40245
291.1325
0
Number of employees
between
275.61
0
(female)
within
28.60904
-2018.93
overall
15.02438
55.32137
1
Total turnover
between
49.64795
1
within
18.11802
-502.476
overall
8.473034
34.25253
0
Total turnover (male)
between
30.91564
0
within
11.35697
-410.527
overall
6.611734
27.6204
0
Total turnover (female)
between
24.50675
0
within
9.692457
-368.055
overall
1.533086
1.12395
1
Business volume
between
1.080366
1
within
0.187078
-0.46691
overall
438516.1
2849000
0
Gross monthly pay in
between
2784779
0
Euros
within
444459.7
-4.42E+07
Avg. weekly work
hours

overall
between
within

39.06646

105

3.121292
3.106473
1.39372

0
0
11.73312

Max
50524
48009.33
2694.51
18531
17208.33
1483.069
2416
2416
573.0244
1551
1551
564.1397
1387
865
561.6117
4
4
3.533086
1.81E+08
1.78E+08
4.51E+07
91
91
63.39979

Table 3.2 (cont’d)
Variable
Overtime
Collective agreement
Works council
Profit
Equal opportunity
officer
Promotion of female
employees
Childcare practice
Parental leave contact

Level
overall
between
within
overall
between
within
overall
between
within
overall
between
within
overall
between
within
overall
between
within
overall
between
within
Overall
between
Within

Mean
0.353845
0.520432
0.654906
3.550509
0.24448
0.054193
0.041214
0.149895

Std. Dev.
0.478167
0.442882
0.222864
0.499588
0.479682
0.174052
0.475404
0.464334
0.099004
1.389494
1.33065
0.555315
0.429783
0.395427
0.206112
0.226401
0.199498
0.120723
0.198786
0.185027
0.101004
0.356972
0.324846
0.177188

Min
0
0
-0.31282
0
0
-0.14623
0
0
-0.01176
1
1
0.550509
0
0
-0.42219
0
0
-0.61247
0
0
-0.62545
0
0
-0.51677

Table 3.3: Frequency of establishment changes in childcare practices
Percent of
Percent of
Establishments
Practice Change
Establishments
(Parental Leave
(Childcare)
Contact)
Added between 2002 and 2004
0.75
4.44
Added between 2004 and 2008
1.16
4.58
Dropped between 2002 and
1.43
4.63
2004
Dropped between 2004 and
0.75
4.44
2008
Present in 2002, 2004, and 2008
0.00
0.00
Not present in 2002, 2004, or
96.66
86.35
2008
Note: Includes establishments appearing in all three years only; N=20,638

106

Max
1
1
1.020511
1
1
1.187099
1
1
1.321573
6
6
6.883842
1
1
0.911146
1
1
0.72086
1
1
0.70788
1
1
0.816561

Table 3.4: Breakdown of sample and childcare offerings by industry
Industry
Agriculture/Forestry/Fishing
Mining/
Manufacturing
Electricity/Water supply
Construction
Trade/Hotels/
Restaurants
Transport/
Communication
Financial Intermediation
Real Estate/
Commercial Services
Public Services
Non-Industrial
Organizations/
Private Households
No Industry Given

Number
of Obs.

%
Obs.

Number
of Estab.

771

1.62

588

1.74

1.56

% Offering
Parental
Leave
Contact
4.93

7541

15.89

5993

17.71

4.43

14.34

323
8984

0.68
18.93

268
7475

0.79
22.09

8.98
1.39

22.29
8.44

7702

16.23

5196

15.35

1.79

9.47

1792

3.78

1221

3.61

3.51

12.89

1313

2.77

809

2.39

10.89

40.06

5559

11.71

3661

10.82

5.54

12.93

8772

18.48

5116

15.12

7.58

28.75

2625

5.53

1619

4.78

4.61

13.22

2078

4.38

1900

5.61

0.87

4.38

%
% Offering
Estab. Childcare

3.7.3 Regression Analyses
The between-establishment analysis is shown in Table 3.5. Establishments offering
childcare practices throughout more of the survey years have higher levels of total turnover,
contrary to Hypothesis 1a, which stated that establishments with childcare practices would have
lower overall turnover on average. This relationship also holds for both male and female
turnover examined separately. Parental leave contact, on the other hand, is associated with lower
levels of turnover for men, but not for women or overall. This provides limited support of
Hypothesis 1b.
The within-establishment model is shown in Table 3.6. Several control variables yield
significant effects. Adoption of a collective agreement and increased profit are associated with
lower turnover overall and for men, but not for women. Turnover also tends to be lower over
107

time, with overall and female turnover lower in 2008 and 2004 than in 2002. Male turnover is
also lower in 2008 compared to 2002, but not 2004. Finally, in partial support of Hypothesis 2a,
adoption of a childcare practice is associated with lower turnover overall and for men. For
women, the coefficient failed to reach statistical significance (p=0.07) but trends in the expected
direction. In opposition to Hypothesis 2b, changes in parental leave contact practices are
unrelated to turnover in all three models.
The combined model which simultaneously includes between- and within-establishment
components is shown in Table 3.7. The model shows effects largely consistent with those of the
distinct within- and between-establishment analyses for key variables. One exception is that
whereas the within-establishment model does not find a relationship between a change in
childcare policy and female turnover, such a relationship is present in the combined model. This
provides additional support for Hypothesis 2a. The model is consistent with the above opposite
direction finding regarding Hypothesis 1a. Consistent with the general lack of support for
Hypotheses 1b and 2b, the between-establishment relationship of parental leave contact and male
turnover is not statistically significant in the combined model. Overall, the combined model
suggests that the relationship between childcare and establishment turnover is robust while that
of parental leave contact and turnover is not.
Hypotheses 3a and 3b state that the relationship between childcare practices and parental
leave contact and turnover are stronger for turnover of women than turnover of men. To test
these hypothess, I calculate Z statistics comparing the unstandardized regression coefficients
across the three models. These comparisons are presented in Table 3.8, and are calculated even
when the coefficients themselves were not statistically significantly different from zero as they
may still be different from one another. This test reveals a statistically significant difference in
the between-establishment coefficients of both childcare and parental leave contact between men

108

and women. These differences did not hold up in the combined model, in which no comparison
produced a statistically significant difference.
Thus, Hypotheses 3a and 3b are only partially supported. There are no differences in
within-establishment coefficients. The positive relationship of childcare with establishment
turnover is stronger for women than men. Though a stronger relationship was hypothesized, the
observed direction of the relationship was not. For parental leave contact, the betweenestablishment effect was more negative for men. Yet once again, these differences do not appear
in the combined model.
The within-establishment and combined poisson models presented in Appendix B are
largely consistent with their linear counterparts presented up with respect to key variables. The
single exception is that the poisson fixed-effects (i.e., within-establishment) regression shows a
negative relationship between the adoption of a childcare practice and female turnover whereas
the linear version did not. This provides additional support for Hypothesis 1a and is also
consistent with the linear combined model. The poisson combined model for total turnover failed
to converge, thus its results are not presented in Appendix B. Overall, the similarity of the
poisson models to the linear models suggest that the models presented do not suffer from the
non-normality of the distribution of the dependent variable, and that the standard error clustering
techniques successfully ameliorated problems associated with linear modeling of count data.

109

Table 3.5: Between-establishment regressions
Total Turnover
SE
B
Proportion of employees
female
-0.015
0.817
**
Total employees
0.046
0.001
Gross monthly pay (10,000)
Overtime
Collective agreement
Works council
No equal opportunity policies
Profit
Business volume
Average weekly work hours
Childcare practices
Parental leave contact practices
Female promotion practices
Industry:
Mining/Manufacturing
Industry: Electricity/Water
supply
Industry: Construction
Industry:
Trade/Hotels/Restaurants
Industry:
Transport/Communication
Industry: Financial
Intermediation
Industry: Real
Estate/Commercial Services
Industry: Public Services
Industry: Non-Industrial Orgs
Constant
N: Observations
N: Groups
*

**

-0.06
-0.421
**
1.887

0.003
0.489

-0.574
**
-2.438

0.570

-0.702
0.450

0.206
0.334

0.886

**

-4.299

**

Female Turnover
SE
B
**

0.486

4.285

0.001

0.028

0.002
-0.174
**
1.123

0.002
0.291

0.002
0.267

0.294

-0.06
-0.285
**
0.733

0.374
**
-1.808

0.339

-1.011

0.311

0.529

0.486

0.123
0.198

-0.623
**
-0.365
0.325

0.004

0.721

0.006
**
5.897

0.569

0.024

0.523

0.695

-0.897

0.639

1.077

-4.790

1.899

-5.742

1.108

-3.593

1.071

-3.992

1.187

-4.548

1.276

-4.071

0.998
1.082

0.018

**

*

**

**

**

-0.003
**
7.571

0.007

-0.306
0.163
*
-0.009

1.212

1.884

-1.378
*
-2.813

0.956

-1.504

1.167

-1.906

1.807

-5.546

3.184

-5.422

1.859

-5.967

1.796

-5.191

1.989

-4.874

2.144

-7.022

-1.948
**
-9.609

1.821

1.086

1.870

-1.362
**
-5.616

1.114

-0.635
**
-4.123

-7.164

1.977

-3.415

1.178

-3.723

0.817

14.126
2.353
35,781
22,798

-10.088
-10.982
-9.462
-9.089
-9.233

**
**

**
**
**

-10.825

-0.015

**

0.495

Male Turnover
SE
B

**

**
**

36,308
23,097

**
**
**
**
**
**
**
**
**

**
*

0.004

3.692

Notes: p≤0.05; p≤0.01
State and year controls not shown in table but were included in all analyses.

110

**
**
**
**
**
**

**

0.446
0.001

0.270

0.113
0.182
0.662

0.990
1.745
1.018
0.984
1.091
1.172

1.024

2.163
35,783
22,799

Table 3.6: Within-establishment regression models with establishment fixed effects
Total Turnover
Male Turnover
Female Turnover
Β
SE
Β
SE
Β
SE
Proportion of employees
0.556
1.458
-0.730
0.801
1.230
0.787
female
Gross monthly pay (10,000)
0.001
0.040
0.001
0.035
0.001
0.011
Overtime
0.176
0.280
0.064
0.144
0.085
0.198
*
*
Collective agreement
0.443
0.305
-0.325
0.193
-0.940
-0.608
Works council
1.179
1.307
1.052
0.972
0.163
0.530
No equal opportunity policies
-1.324
0.880
-0.873
0.515
-0.395
0.544
*
Profit
-0.201
0.166
0.096
-0.002
0.110
-0.206
Business volume
Average weekly work hours
Total Employees
Childcare practices
Parental leave contact
practices
Female promotion practices
Year: 2004
Year: 2008
Constant
N: Observations
N: Groups
*

Notes: p≤0.05;

**

0.899
0.003
0.023
2.176

1.967
-0.002
-0.001
-1.871

0.674
0.002
0.013
1.059

-2.782

-0.219

1.228

-0.500

0.554

0.328

0.880

1.055

1.650
0.312

0.679
-0.065

1.083
0.175

0.418

0.804
0.213

0.391
-1.182
1.651
4.106
37,793
23,602

*

-4.489

-0.739

*

*

**

**

p≤0.01

111

0.642
0.002
0.026

**

0.350
0.002
0.015
1.40

2.599
0.000
0.026

*

0.204
-0.428
-0.034
2.145
37,262
23,305

-0.662

**
*

0.298
-0.760
1.895
2.686
37,264
23,306

Table 3.7: Combined linear model
Variable

Total Turnover

Between-Establishment

Within-Establishment

Β
Proportion of employees
female
Gross monthly pay (10,000)
Overtime
Collective agreement
Works council
No equal opportunity
policies
Profit
Business volume
Average weekly work hours
Childcare practices
Parental leave contact
Female promotion practices
Total Employees
Proportion of employees
female
Gross monthly pay (10,000)
Overtime
Collective agreement
Works council
No equal opportunity
policies
Profit
Business volume
Average weekly work hours
Childcare practices
Parental leave contact
Female promotion practices
Total Employees
Constant
N: Observations
N: Groups
*

**

SE

Male Turnover
Β

SE

Female
Turnover
Β
SE

0.669
0.003
0.240
*
-1.009
1.383

1.636
0.043
0.300

-1.035
0.003
0.119

0.904
0.034
0.152

1.621
0.000
0.099

0.884
0.011
0.209

0.471
1.315

-0.619
1.233

0.322
0.971

-0.385
0.210

0.210
0.528

-1.378

0.885

-0.959

0.533

-0.371

0.524

-0.229

0.174

-0.220

0.101

-0.022

0.113

2.02
-0.001
*
-5.256
-0.362
1.424
0.025

0.760
0.005

0.367
0.002
*
-3.261
-0.566
0.883
0.024

0.271
0.003

0.608
0.003

1.372
0.554
1.097
0.014

1.624
-0.003
*
-2.137
0.263
0.608
0.002

-0.789

1.761

0.969

2.565

-0.063
-0.528
**
2.938
-1.700

0.051
0.542

0.039
0.309

-0.061
-0.270
**
1.258
-1.055

0.018
0.322

0.608
1.610

-0.002
-0.252
**
1.671
-0.767

-1.332

2.096

-1.090

1.301

1.036

-0.411

0.237

-0.045

0.140

-0.343
*
-0.334

-1.602

0.914

-0.192
*
-0.010

0.411

-1.336

0.645

0.004

0.011

**

2.108
1.205
1.649
0.022

-0.000 0.007
**
14.245
3.968
-1.269 2.228
-5.369 2.872
0.022 0.024
*
11.246
4.676
36,308
23,097

-3.235

*

**

*

0.378
1.081

5.682
2.191
-1.201 1.301
-3.499 1.835
-0.005 0.015
**
10.309
3.047
35,781
22,798

**

**
**

*

**
**

8.993
-0.265
-2.014
0.027

1.014
0.849
0.787
0.012
0.978

0.330
0.810

0.143
0.004
2.354
1.223
1.511
0.015

1.089 2.284
35,783
22,799

Notes: p≤0.05; p≤0.01
Fixed effect controls for one digit industry, German federal state, and year are included
but not presented.
112

Parental Leave
Contact

Childcare

Table 3.8: Comparison of undstandardized regression coefficients for male and female turnover
Separate Models
Male Turnover Equation
Female Turnover Equation
Practice
Level
Z
Coefficient
Coefficient
Coefficient Coefficient SE
SE
**

Between

1.884

0.721

5.897

0.662

Within

-2.782

1.401

-1.871

1.059

-0.519

Between

-1.504

0.569

0.024

0.523

-1.977

Within

-0.500

0.554

0.328

0.880

-0.796

-4.100

*

Combined Models
Level

Parental Leave
Contact

Practice

Childcare

Male Turnover Equation

*

Female Turnover Equation
Z

Coefficient

Coefficient
SE

Coefficient

Coefficient SE

Between

5.682

2.191

8.993

2.254

-1.053

Within

-3.261

1.372

-2.137

1.014

-0.659

Between

-1.201

1.301

-0.265

1.223

-0.524

Within

-0.566

0.554

0.263

0.849

-0.818

Notes: p≤0.05; ** p≤0.01;

**

p≤0.001

113

3.8

Discussion
The results are somewhat surprising in that, contrary to the predictions of strategic HRM

theory and Hypotheses 1a and 1b, establishments with childcare practices did not have lower
turnover than those that without childcare practices. In fact, those with a higher prevalence of
onsite childcare reported higher turnover. There are several possible explanations for this
unexpected result.
One possible though unlikely interpretation of this finding is that childcare WFPs cause
higher turnover. This could be true if establishments with childcare practices attract employees
that are more likely to turnover. For example, this result might occur if establishments with
childcare practices attract more employees with children who plan to exit the workforce in the
future. This interpretation is unlikely to be accurate as it is inconsistent with strategic HRM
theory and the within-establishment and combined analyses. Another more plausible explanation
is that establishments with the highest levels of turnover are most likely to adopt the childcare
practices in the first place, in order to drop their presumably problematic turnover rates.
The between-establishment analysis, typical of most strategic HRM studies, does not
provide sufficient empirical justification for either argument on its own. The latter explanation is
supported by the within-establishment analyses. These analyses show that in the years
establishments do have childcare practices in place, they tend to have significantly lower levels
of turnover compared to the years in which they do not. Together, these two effects provide a bit
more insight into the nature of the relationship between childcare practices and turnover.
The between-establishment coefficients are therefore best interpreted not as the effect of
a childcare practice on turnover but rather as the relationship between turnover and childcare
offerings. In other words, it tells us which establishments offer childcare but not necessarily
whether childcare is effective at reducing turnover. The results suggest that it is high-turnover
114

establishments that are implementing childcare practices. This makes sense intuitively as
establishments with turnover problems are likely to be looking for solutions. Furthermore, the
larger coefficient for female turnover, in partial support of Hypothesis 3a, suggests that
establishments with female turnover problems are even more likely to adopt childcare practices.
These results also suggest that childcare practices might be offered more typically on a needs
basis rather than as a reward for high performing employees or because the organization itself is
performing better. This may not be the case for other strategic HRM practices. There is at least
some evidence to suggest that the causal relationship between practice and organizational
performance may vary from one practice to the next (Van Iddekinge et al., 2009).
As such, the practice of continued contact during parental leave exhibits a very different
relationship with establishment turnover, emphasizing the importance of examining practices
separately. While there are no within-establishment relationships, and although the relationship
is not robust to the combined model, there is a negative between-establishment relationship with
turnover of males. This suggests that establishments with lower male turnover may be more
likely to offer parental leave contact. This practice may be more attractive to men because they
are generally more likely to return to work after parental leave than women, and for men who
wish to use part of the three total years of unpaid parental leave available between parents, this
practice makes a return to work easier. An easier return to work is also better for the
establishment because it can prevent the need for bringing returning employees up to speed with
information they may have otherwise missed.
However, the lack of within-establishment effect suggests that this practice may not be as
important to employees as childcare practices. Changes in this practice were not associated with
changes in turnover levels. This pattern of results also suggests that the low turnover of men may

115

be driving the adoption of the policy rather than the other way around. Also of note is that even
though the between-establishment relationship was statistically significant for men and differed
between men and women, these relationships were not robust to the combined model.
The within-establishment coefficients are better suited to telling us what WFPs do in a
given establishment. They represent a natural experiment in which we observe a number of
changes in practices and turnover levels within an establishment. While the number of practice
changes is relatively small, as shown in Table 3.3, the large sample provides sufficient incidence
of these changes to test their relationship to turnover. The within-establishment coefficients
support strategic HRM theory in that the adoption of childcare practices are associated with
reduced turnover.
The lack of a statistically significant difference in the WFP coefficients for the withinestablishment male and female turnover models suggest that WFPs are similarly beneficial for
both men and women, even if they appear to be adopted more commonly when female turnover
is higher. This finding may be in part due to the erosion of traditional gender roles. For dual
earner couples, it may not matter much which spouse’s establishment offers childcare practices
as both partners benefit. A woman may benefit from having a partner who works in a childcare
offering establishment just as much as if her establishment did as either way childcare is
provided, allowing both parents to work full time without constraint. Similarly, men may benefit
from childcare practices just as much as women as they begin to take on more caregiving
responsibilities on average.
Together, the results of the three models suggest that childcare practices may be able to
reduce the German childcare gap in a way mutually beneficial to both employees and employers.
Furthermore, the use of these practices may be especially helpful in keeping women in the

116

workforce who might otherwise use the Betreuungsgeld monthly childcare allowance to remain
at home longer after a child’s birth. If Betreuungsgeld’s critics are correct that it will dissuade
mothers from rejoining the labor force after childbirth, then establishment childcare practices
will likely become even more effective turnover deterrents for those on the margin. For lowwage mothers who would not lose substantial income from taking Betreuungsgeld instead of
working, having access to a reasonable establishment-driven childcare option may make
continuing work a competitive option.
The issue of Betreuungsgeld and even the existence of a childcare gap in Germany
highlight the importance of considering the public policy context in evaluating the strategic
plausibility of HRM practices (Piszczek & Berg, forthcoming). Even though there is a substantial
childcare gap in Germany, other Western countries have even less institutional support for
childcare. In the United States, maternal leave is limited to 12 unpaid weeks and there are no
public daycare options for young children. In such countries with weaker childcare support,
childcare practices may be even more strategically effective in reducing turnover and attracting
applicants, as there is a larger potential labor market with unmet needs. On the other hand, the
current high levels of unemployment in the United States make turnover and applicants less of a
problem to begin with. Based on the current study, establishments thus may be even less likely to
adopt childcare practices in the U.S. compared to Germany as turnover may be less problematic,
but they may find them to be more effective when they do.
Country-level economic factors are also tied to these decisions and may impact the
results. Countries in poorer economic conditions, whether because of economic downturn or
more permanent reasons, may have trouble offering expensive public childcare options. In these
cases, employers may not necessarily be willing to pay the cost to offer them either. Similarly,

117

when the labor force is economically stronger, employees can afford private daycare more easily
and the impact of establishment-sponsored childcare may not be as strong. In a particularly weak
labor market, the effects of establishment-sponsored childcare may also be weaker as employees
will be less likely to quit their jobs regardless of childcare options as the prospect of finding
alternative employment might be more difficult. Overall, further research is needed to understand
how the economic climate may impact the relationship between establishment work-family
practices, public work-family policy, and employee turnover.
3.9

Limitations and Conclusions
The current study has a number of limitations. Most notably, it does not use individual

level data though a number of non-organizational, individual factors might influence turnover at
the individual level. The sampling frame helps to eliminate this issue. Though it is not possible to
randomly assign childcare practices to an establishment, the stratified random sample of
establishments helps ensure that individual-level causes for turnover are distributed evenly
across childcare and non-childcare establishment-year observations. Similarly, reason for
turnover cannot be examined. It may be the case that childcare practices affect an employee’s
performance and therefore likelihood of involuntary turnover, but involuntary and voluntary
turnover cannot be differentiated in the present study. A second limitation is the surveys used
rely on information from a single informant whose knowledge of organizational practices may
not be complete. Though establishments were instructed to have a knowledgeable person provide
the information and should have given the survey’s ties to government, taxes, and social security,
not all may have done so. Finally, the values for turnover come from only the first half of the
year due to the timing of the survey. Though I control for several aspects of establishment
performance, large-scale turnover events such as layoffs or seasonal employment may have
118

influenced the turnover counts or resulted in major fluctuations in turnover during the second
half of the year which cannot be accounted for in the data.
Despite these shortcomings, this study makes several contributions. First, it provides a
more nuanced and more generalizable analysis of the relationship between establishment
childcare practices and turnover of men and women in Germany. The study provides evidence
for the strategic value of childcare practices in the context of Germany, as the results suggest that
childcare practices are associated with reduced turnover levels in establishments with higher
average levels of turnover. This argument is consistent with strategic human resource
management theory. This study also provides a contribution to the ongoing debate of the causal
relationship between strategic human resource management practices and organizational
performance. By parsing the analysis into within-and-between establishment components and
including establishment fixed effects, I show that the reverse-causation argument that childcare
practices are adopted by high-performing organizations is not supported. On the contrary,
childcare practices are adopted by organizations with typically higher levels of turnover.
Finally, the study answers the call for more contextualized international work-family
research by providing an in-depth discussion of how environmental factors such as public policy,
labor market demands, and family culture interact to create strategic opportunities for employers.
Future research should seek to identify other country contexts with similar or different
environments regarding childcare to further examine whether the context affects the strategic
effectiveness of childcare policies. Future research should also seek to identify whether other
types of WFPs might hold strategic value in Germany and similar contexts.

119

4
4.1

An Integrated Organizational Work-Family Response Framework
Introduction
In Chapter 1, I introduce three frameworks describing the organizational work-family

response developed in the 1990s: institutional pressure, rational choice, and agency. I discuss the
strengths and weaknesses of each framework in the context of current research. The institutional
pressure framework (Goodstein, 1994, 1995; Oliver, 1991) provides insights about why
organizations choose to adopt work-family policies and practices or not, but does not discuss
how policies are chosen or implemented. The rational choice framework (Glass & Estes, 1997;
Osterman, 1995; Seyler et al., 1995) emphasizes the importance of business outcomes, but does
not say much about what factors go into the decision of whether a particular work-family policy
or practice will be cost effective. Finally, the agency perspective (Kirchmeyer, 1995) considers
the ability of the organization to choose the specific policies and practices they adopt and their
different meanings for employees, but does not discuss how the decision to adopt them is made
or how they are implemented.
In order to have stronger theoretical grounding at the organization level, researchers can
borrow ideas from these frameworks, update them, and integrate them with knowledge of the
work-family interface gained from more current research. Research with stronger organizationlevel theory can help explain complexities in work-family policies and practices (e.g. uneven
application of policies across different employee groups) and raise new research questions about
how work-family policies and practices are chosen, implemented, and related to organizational
success. Organization level theory will also help answer scholars’ call for more research
supporting the business case for work-family practices (Kossek, Baltes, et al., 2011; Kossek &

120

Friede, 2005) by helping explain how and when organizations can successfully implement workfamily policies and practices.
The goal of this concluding chapter is to integrate the three organizational work-family
response frameworks discussed in Chapter 1 with current literature in order to generate a more
useful organization-level theoretical framework that will further both individual and
organizational work-family research. First, I will discuss the previous two chapters of this
dissertation in the context of the organizational work-family response and as evidence for the
need of an updated, integrated organization-level framework. Second, I discuss the benefits of an
updated framework and why it is critical to have theoretical foundations for organization-level
work-family policy decisions for the continued growth of both the organization and individual
work-family literatures. Third, I will introduce the integrated framework and describe it in depth,
specifying precisely how it integrates the strengths of previous frameworks with current workfamily knowledge. Fourth, I discuss how the new integrated framework can be used in future
research, including areas it highlights as research gaps and potential questions it raises for further
investigation. Fifth and finally I offer some concluding and summary thoughts about this
dissertation as a whole.
4.2

Chapters 2 and 3 and the Organizational Work-family Response
The previous two chapters, much like other current studies with an organizational

component, contribute to the goal of developing a new organizational work-family response
framework. Each chapter stands on its own as a distinct empirical study, but each also
contributes theoretically to a more complete organizational work-family response framework that
can explain complex patterns of work-family policy and practice adoption and implementation
within and across organizations. Both chapters’ studies address ways that organizations respond
121

to their employees' work-family needs. Each study makes its own empirical and theoretical
contributions to specific aspects of this response, thus also contributing to narrower work-family
literatures. Each study also shows that previous frameworks require updating to be useful to and
consistent with current work-family research. The study presented in Chapter 2 demonstrates the
need to account for organizations that respond to employees’ competing role demands by
attempting to control their non-work roles. Similarly, Chapter 3 shows the importance of
considering the broader environment and related pressures when examining the adoption and
effects of work-family practices. I begin by summarizing each of the two empirical studies, and
then discuss in greater depth how they demonstrate the need for an integrated, multi-stage
organizational work-family response framework.
Chapter 2 is a study of organizational pressure to use interactive communicative
technology to perform work role tasks while physically in the family domain. As technology
continues to become more efficient and more portable, research on this topic is becoming more
important. Yet, the role of technology is under-researched in the work-family domain. This
chapter discusses the research showing how information and communicative technologies can
have conflicting outcomes. They may either help employees work with more physical and
temporal flexibility or contribute to more time spent working and invasion of work into other life
domains. Only recently have these perspectives been combined, typically using the job demandsresources perspective (e.g., Derks & Bakker, 2010), and few studies have examined the part of
the organization in setting expectations around after-hours technology use in the form of afterhours electronic communications expectations (Fender, 2010; Fenner & Renn, 2010).
The study builds on recent research utilizing the job demands-resources perspective to
shed some light on the conflicting observed effects of use of information and communicative

122

technologies for work purposes while at home. The study adds theory to explain the mechanisms
through which technology can function as either a job demand (firm technical control) (R.
Edwards, 1979) or job resource (boundary theory) (Nippert-Eng, 1996). Using these theoretical
explanations, the study argues that one key determinant in whether technology is experienced as
a demand or resource is organizational pressure to use it in certain ways, i.e. for work purposes
after normal working hours.
Importantly, the study also explains that this pressure might not affect the experience of
information and communicative technology use equally for all employees. Thus, it considers
how this pressure is experienced by individuals with different work-family role management
preferences. Findings support an overall positive relationship between organizational
expectations for after-hours technology-based work and employee emotional exhaustion, which
is subsequently linked to turnover intentions. However, this relationship is largely driven by
segmenters, the employees who wish to keep work and family domains distinct in their lives.
Integrators, i.e. those who prefer to amalgamate their work and family roles, are largely
unaffected by these pressures. Overall, this study contributes to the small but growing body of
empirical research explaining the conflicting potential effects of after-hours technology use and
sheds some light into the question of when and why technology use may have either desirable or
undesirable effects on employee well-being. Finally, it highlights the importance of
organizational policy around after-hours electronic communications.
Chapter 3 also makes substantial contributions to its more narrow literature of workfamily practices as components of a strategic human resource management system. This study
adds to the body of research supporting the empirical link between adoption of a work-family
practice and desirable organizational outcomes, specifically demonstrating a link between the

123

adoption of a childcare practice and reduced turnover at the organization level. It contributes to
the limited body of research examining the relationships between organizational turnover and the
specific work-family practices of organization-sponsored childcare and parental leave contact.
The study highlights the importance of context by discussing at length the reasons that
childcare and parental leave contact may harbor strategic advantage. It reviews German public
policy surrounding early childhood daycare and education in order to demonstrate its
undersupply for the current German labor market, which is expected to include more women in
the future. This can help explain why research sometimes finds work-family policies to be linked
to desirable organizational results, but other times fails to find this link; there must be sufficient
demand for a practice in order for it to hold strategic value. In this case, the demand is in part due
to the lack of public options surrounding childcare.
Another important contribution of Chapter 3's study is the analytical strategy
accommodated by the large multi-year dataset. Because multiple observations were available for
each establishment in the sample, variation in turnover can be partitioned into two components.
Thus, work-family policies are linked to variation in turnover between different establishments
and to variation in turnover within an establishment as it changes its practices over time. An
interesting pattern reveals itself; while the adoption of a childcare practice is associated with a
reduction in turnover within an establishment, establishments with higher levels of turnover are
more likely to have such practices. By analyzing each variance component separately, a much
stronger argument can be made as to the causal direction of the relationship between practice and
performance. This has been an area of controversy in strategic human resource management
research (Wright et al., 2005). This pattern of results suggests that while childcare policies are
effective in reducing turnover, they are necessarily not being adopted to strategically reduce

124

turnover. Instead, they are more likely to be adopted in order to address problematic levels of
turnover.
Each study can therefore stand on its own empirical and theoretical merit, contributing to
narrower work-family research literatures. Despite this, both studies have one important thing in
common: they are each concerned with a way an organization can respond to the work-family
needs of employees which is not explained by existing organizational work-family response
frameworks or current individual-level research. In order for a framework to be useful, it must be
able to explain these complex patterns of work-family policy and practice adoption and
implementation.
Though electronic tethering to an organization has received individual-level research
attention as technology has continued to evolve (e.g., Renaud et al., 2006; Richardson &
Thompson, 2012; Turkle, 2007), the organization-level component of Chapter 2’s study (afterhours electronic communications expectations) has only recently received research attention,
largely in terms of conceptual development (Fender, 2010). This type of general response to
employees’ work-family needs—manipulation of the non-work domain—is unaccounted for in
any existing organization-level work-family frameworks.
Given the results of Chapter 2 that show its importance in influencing individual
outcomes such as work-family conflict, emotional exhaustion, and turnover intentions, it is
critical for researchers to understand the antecedents to the adoption of such a practice. Why
would an organization seek to manipulate its employees’ working time? What are the conditions
that allow such manipulation to manifest and remain unchallenged? Are there other options
under such conditions that may be mutually beneficial to both the organization and its
employees? An integrated multistage framework that includes this type of organizational practice

125

and links it to environmental characteristics and policy implementation can help answer some of
these questions by elucidating why employers might not offer work-family practices or policies
or pressure employees to use their non-work time for work purposes.
Chapter 3 also supports the need for an organization-level integrated framework. It shows
that the rationale for adopting a particular work-family practice can be the result of a rich
interaction of contextual factors. Recent research suggests that these contextual factors are
frequently simplified or ignored in research, resulting in the misattribution of policy effects and
misinterpretation of study results (Piszczek & Berg, forthcoming).
Chapter 3 is grounded in a strategic human resource management perspective, and while
it does show that the adoption of childcare practices is associated with lower turnover, it also
shows that the adoption of childcare practices is associated with high-turnover establishments.
This is most consistent not with a strategically motivated policy adoption but rather a labor force
pressure based policy adoption. Further complicating the matter is that the lack of public policy
surrounding childcare increases the labor force pressure on organizations to adopt them instead.
These contextual nuances in the adoption of a work-family policy or practice may affect the way
a policy or practice is implemented, but such nuances are not accounted for in existing
organizational work-family response frameworks and are largely absent in most individual and
multi-level studies (Piszczek & Berg, forthcoming).
4.3

Benefits of an Updated Framework
There are numerous advantages to revitalizing the organization-level work-family

response framework. Most importantly, a new framework that integrates the key concepts from
previous frameworks with newer knowledge of the work-family interface can be useful for
grounding and contextualizing future organizationally focused work-family research. It can also
126

drive new research by clarifying links between organizational and individual level phenomenon
and identifying organizational and environmental contingencies in the effectiveness of workfamily policies and practices. It raises questions about how work-family policies are chosen and
implemented within and across organizations, what organizational and institutional contextual
factors impact their effectiveness, and how best to connect a work-family policy or practice to
the goals of the organization. As one study explains, “Without a generalizable theory of
contingent response to pressures for work-family programs, the field is left with isolated findings
that often seem to contradict each other” (Ingram & Simons, 1995). However, little progress has
been made in this regard in the last 20 years. As an example, one recent study of an
organizational work-family intervention sought to customize the intervention for each branch of
the target organization, but noted that "adaptation across sites does not fit neatly into scientific
paradigms" (Kossek et al., 2014). Few studies have examined the work-family response as an
organization-level phenomenon, and those that have largely characterized it as response to
institutional pressure without accounting for organizational strategy, practice implementation, or
even different types of practices.
Instead, work-family research has focused on the effects of particular practices on
individual employees. By focusing on the effects of work-family policies and practices at
primarily the employee level, researchers have struggled to classify them into a meaningful
system that can be easily and consistently applied to research and practice at the organization
level. Organizational practices are typically classified not by their purpose but rather by
functional ways they affect employees. While there is value to organizing work-family practices
in this way, such as by whether a practice grants flexibility in timing of work, place of work,
amount of work, or short-term breaks in work (Kossek & Michel, 2011), or as encouraging

127

integrating or separating (Rau & Hyland, 2002; Rothbard et al., 2005), this value lies primarily at
understanding how they affect individual employees. There is also value in considering the
organizational rationale for the adoption in order to better define and predict the success of a
particular practice and whether it will be used as expected by employees.
Grounded at the organizational level, a work-family response framework shifts the focus
from how policies and practices affect employees to how policies and practices fit within the
broader vision of the organization. Viewing a practice in the context of such a framework raises
new questions, as it allows organizations to assess the consistency of that practice with its highlevel human resource management goals and create synergistic packages of practices that reach
them. This answers the call for human resource management research that considers uneven
policy and practice application within organizations and identifies contingencies that impact
effectiveness (Kaufman, 2010).
Similarly, an updated framework will allow practitioners and researchers to compare and
contrast systems of work-family policies and practices across organizations. Understanding
practices with more organization-level context would contribute to studies of work-family
practices grounded in strategic human resource management (e.g., Chapter 3) which are growing
in popularity but struggle to demonstrate when and why particular work-family practices are
strategically viable. In other words, an integrated framework can help put boundary conditions
around the strategic value of work-family policies and practices by identifying the contexts in
which they work best. Similarly, such a framework can help avoid study interpretation problems
due to misattributing practices to an organizational strategy when they may be the result of
another motivation, such as compliance with the law (Piszczek & Berg, forthcoming).

128

As an example, consider an organization that has decided to start a work-family initiative
because it is losing its young up-and-coming engineers to other companies. There are many ways
this initiative could be enacted: flexible work hours, onsite childcare, a work-family supportive
culture, and family-supportive supervisor training are just a few examples that have been
empirically linked to more desirable employee outcomes such as increased perceptions of
organizational work-family support and lower work-family conflict (e.g., Allen, 2001; Frye &
Breaugh, 2004; Hammer, Kossek, Yragui, Bodner, & Hanson, 2009; Rau & Hyland, 2002).
Considering all the potential effects of work-family policies and practices, an often unconsidered
question is which effects are most consistent with the organization’s goals? Furthermore, what is
the most cost-effective way or implementing such an initiative in order to achieve those goals?
If the organization is only concerned about its engineers as key strategic employees,
onsite childcare may be prohibitively expensive if it is implemented broadly such that other
employees can access it as well. At the same time, if only engineers are permitted access it may
create resentment or perceptions of injustice among other employees. A broad work-family
supportive organizational culture may encourage other employees for which there is no turnover
issue to sacrifice their work tasks for family tasks, potentially creating new staffing problems. On
the other hand, flexible work hours for engineers or training engineering team leads or
supervisors to better help the employees manage work-family demands may help reduce turnover
intentions of young talented engineers without affecting other employees. The source of the
pressure to adopt a policy in the first place and the expected results of its adoption play a key
role in how the policy is implemented, but current research does not consider these important
organization-level factors.

129

Contrast the previous example with one of a similar company also planning a workfamily initiative, but for the purpose of putting forth a family-friendly image to customers rather
than retention of an employee subgroup. Customers may perceive numerous types of policies and
practices as family-friendly, but some are more visible and easily publicized than others. The
general public may not hear much about family-friendly supervisor training, but an organization
may find more formalized practices such as shorter working hours easier to publicize and help
impart desired organizational legitimacy on the public. Organizations such as Chick-fil-A and
Hobby Lobby are well-known for being closed on Sundays to allow employees time for their
non-work roles. Similarly, as more retailers have begun to open their doors on the Thanksgiving
holiday, many organizations such as Costco, Nordstrom, and Menard’s refuse, publically
declaring that they want their employees to be able to celebrate the holidays with their families
(Davis, 2013).
As another example, consider the practice of flexible work hours. Studies of flexible
work hours have linked it to different types of organizational response. Goodstein (1994) argues
that its presence (without childcare practices) is evidence of a broader “avoidance” response
while Rau and Hyland (2002) suggest it supports a broader “integration” response. In fact,
flexible work hours can support a variety of organizational work-family goals. If this practice is
adopted as a retention strategy, then it should be linked to reduced turnover. If it is adopted as an
attraction strategy, it is more important to link it to an increase in job applicants regardless of its
effect on turnover. If it is adopted as part of a corporate social responsibility initiative, the key
outcome is employee well-being or perceptions of organizational legitimacy, and organizational
performance is a less proximal outcome. If it is part of a marketing campaign to appeal to
consumers then sales are more important. Even though flexible work hours are linked to a variety

130

of desirable outcomes, if an organization's implementation of its chosen strategic response
includes flexible work hours, the rationale behind the adoption of that strategic response
determines which outcome defines the practice’s success from an organization’s perspective.
Briefly stated, a successfully implemented practice addresses the pressure that led to its adoption
in the first place, but there is no way to define success without knowing the goal and thus the
pressure.
Another advantage to an updated framework is the resulting clarity of theoretical
mechanisms linking organizational practices to relevant individual outcomes, as research
suggests that the motivation behind a work-family practice’s formulation plays a key role in how
it is implemented (Piszczek & Berg, forthcoming). An integrated framework can link the entire
chain: the initial pressure to adopt work-family practices, the policies chosen to respond to the
pressure, the implementation of the policies, and their expected organizational and employee
outcomes. Such a framework is holistic in that it is based on a multi-stage perspective of
organizational work-family policy adoption and implementation.
Additionally, most work-family role management theories are focused on the individual
level (e.g. boundary theory, border theory) (S. C. Clark, 2000; Nippert-Eng, 1996). A framework
that includes both the conditions that lead to the adoption of a work-family policy or practice as
well as the different ways they might be implemented would help link organization-level
phenomena into these individual level theories by clarifying which individual-level mechanisms
organizations should be seeking to activate through their policies and practices. Identifying the
links and mechanisms address recent calls for research and theory bridging the macro and micro
human resource management literatures (Becker & Huselid, 2006; Huselid & Becker, 2010).

131

Finally, such a framework would help to more richly contextualize an organization’s
work-family policies and practices. This would make for easier comparisons among different
systems of work-family policies and practices across departments, organizations, and countries.
Similarly, it would make contextualization of a study or a particular type of work-family policy
or practice easier and produce more accurate generalizations of a study’s results to other
organizations. Given the importance of contextual factors such as the public policy environment
in the implementation of organizational work-family practices and policies (Piszczek & Berg,
forthcoming; also see Chapter 3), having a framework to organize organizational policies and
practices that includes these types of pressures facilitates comparison across them and responds
to critiques of the lack of institutional consideration in human resource management literature
(Kaufman, 2010). The framework could also be used to compare practices on more than their
content; for example, it would facilitate comparisons based on rationale for adoption,
organizational goal, definitions of success, consistency or method of implementation, and
accessibility.
These theoretical contributions also benefit human resource managers involved with
strategic decisions. An integrated framework can help practitioners make better decisions about
how to deal with the need for a work-family response. As the business case for work-family
policies and practices remains unclear due to mixed empirical links to organizational business
outcomes and a lack of identified organization-level contingencies in policy effectiveness
(Beauregard & Henry, 2009; Bloom et al., 2010; Konrad & Mangel, 2000; B. Lee & DeVoe,
2012; Perry-smith & Blum, 2000), practitioners would benefit from additional guidance in
determining which types of practices and policies might best suit their needs, and how those
practices and policies should be implemented. An integrated multi-stage framework would allow

132

practitioners to more easily identify the institutional pressures they face, what policy strategy
they should pursue, and how to implement that strategy. Without an integrated framework, these
links are scattered and unconnected in research, making work-family policy decisions difficult.
Considering these arguments and examples, the benefits of an integrated organizational
response framework are clearer. There are many ways a particular organizational work-family
policy or practice might be implemented and specific practices and policies may be compatible
with multiple organizational goals. Researchers need a more intricate, multi-stage framework to
better distinguish between the different forms of the organizational work-family response and
what practices and polices (or lack thereof) are consistent with such a response. Researchers also
need a framework to systematically link organization-level strategy to actual policies and
practices. Current frameworks are both too broad and too narrow to accomplish this task. They
are too narrow in that a contextualized view of the organizational work-family response is absent
in any single framework, and they are too broad in that they do not distinguish between the many
ways a chosen response can be enacted and which organizational goals such a response might
best serve.
4.4

An Integrated Organizational Work-family Response Framework
Figure 4.1 presents the updated, integrated organizational work-family response

framework. This expanded framework brings the previous work-family response frameworks in
line with current research and theory, including insights gained from Chapters 2 and 3 of this
dissertation, and integrates key components of the institutional pressure, rational choice, and
agency frameworks. The updated framework links the critical concepts from each framework
and updates them to be consistent with current research. Importantly, the framework outlines the
process of determining whether or not to adopt a particular policy or practice, and what form that
133

policy or practice should take. The framework therefore accounts for an organization’s response
to work-family pressures but can be revisited any time to consider new specific policies and
practices when pressures change or to create synergistic systems of work-family policies and
practices.
The framework takes the form of a multi-stage model, and is split into four main
components, represented by the large boxes. The top-most box consists of environmental
components which contribute to the pressure to adopt work-family practices an organization
experiences. The second box is policy typing, in which organizations determine precisely which
practice or policy might best meet their goals and thus warrant consideration for adoption. The
third box consists of the strategic value an organization expects from that policy or practice. The
bottom-most box is the implementation of that policy or practice. I next discuss each box, its
development, and its link to previous and current literatures.
The "environment" box considers the broader institutional environment within which an
organization operates. The box is informed by the main concept of the institutional pressure
perspective, namely that organizations are responding to the general idea that employees have
multiple roles. Institutional pressures can come in several forms; Oliver (1991) described five
institutional factors defined in Chapter 1: cause, constituents, content, control, and context.
These do not translate well to current research, particularly to the work-family domain.
Goodstein (1994) used organizational size, proportion of female and parent employees, public
sector status, prevalence of policies in a sector, and extent of sector professional organizations to
respectively operationalize these five sources of institutional pressure. Such factors may
contribute to an organization's felt institutional pressure, but more broadly considering more

134

recent research, public policy and labor market demands appear to be the key institutional forces
contributing to institutional pressure.
Public policy pressures are higher when there are more regulative institutions requiring
organizations offer particular work-family practices. Organizations in the United States
experience less of this pressure than other Westernized countries. For example, the United States
has no laws mandating paid vacation, which are common in Europe. Lack of adherence to these
regulative mandates is often punishable by fines and restitution for affected employees, creating
a financial incentive to comply. Public backlash is also possible and potentially damaging for
organizations that do not comply with these mandates. Importantly, these policies and practices
can also vary at the local level. More and more large cities in the United States are passing paid
sick leave mandates, creating even more localized policy variation. Such public policies can be
key drivers of organizational policy and practice adoption and set minimum standards for an
organization to distinguish itself from another through these policies and practices (Piszczek &
Berg, forthcoming).
Labor market demands for the adoption of work-family policies and practice are higher
when an organization's labor pool places high value on such practices. This is best exemplified
as a drive to create person-environment fit among employees, which is predictive of desirable
outcomes such as lower role conflict (Z. Chen et al., 2009; Kreiner, 2006). As younger workers
tend to place more value on control over their work-family boundaries (Burke, 2004; Cleveland
& McCarthy, 2013), organizations seeking young talent may feel more pressure to adopt workfamily practices. On the other hand, organizations that draw from a broader labor pool, or
perhaps have the benefit of an excess supply in the labor market, experience less pressure. These
organizations should be able to find the employees they seek without consideration of work-

135

family practices and policies. Additionally, employee unions can and do seek provisions
providing work-family practices and policies through collective bargaining (Berg et al., 2013;
Berg, Kossek, Misra, & Belman, 2014; Berg & Piszczek, 2014; Glass & Fujimoto, 1995; Seyler
et al., 1995).
Labor market demands also create pressure because they affect the work-family practice
offerings of other organizations, including those competing for talent. Generating sustainable
competitive advantage is a primary goal of strategic human resource management; with regard to
work-family policies and practices, researchers generally agree that the key area in which this is
possible is through the recruitment and retention of employees (Berg et al., 2003; Casper &
Buffardi, 2004; Glass & Estes, 1997; Osterman, 1995; Rau & Hyland, 2002). Thus organizations
that wish to compete on these dimensions experience pressure through the policies and practices
of their competitors. In sectors where work-family policies and practices are generally not
offered, there will be much lower pressure on any one organization and relatively less extensive
policies and practices can hold strategic value. On the other hand, if organizations within an
institutional environment commonly offer some practices, potential applicants may grow to
expect them and thus the pressure to offer them rises. Through benchmarking, organizations will
ensure that they do not fall too far behind their competitors in offering human resource
management practices that might help attract or retain employees. For example, though vacation
leave is not legally mandated in the United States, many employees expect to be able to take a
temporary leave from work from time to time and this practice is commonly offered by
organizations.
Pressure to adopt work-family policies and practices can thus be explained at a general
level with these two variables. Complicating things, however, is that these pressures are closely

136

interrelated; labor market demands for certain work-family policies and practices define what
policies and practice have competitive potential, and public policy creates a baseline from which
organizations must determine what practices might be able to differentiate them from
competitors (Piszczek & Berg, forthcoming). When an organization is considering the adoption
of a work-family policy or practice, the exact nature of this experienced pressure will play an
important role in determining which types of policies and practices might best align with
organizational goals. For example, in an institutional environment where public policy mandates
that organizations offer one year paid maternity leave, an organization may be unlikely to adopt
an additional one month of paid maternity leave as it may be insufficient to differentiate itself
from other organizations. In another context where there is no mandated maternity leave at all,
such as in the United States, one month paid maternity leave may have enough of an impact to
bring additional attention to the organization as a choice workplace considerate of its employees'
family role demands.
This leads to the second box: policy typing. Typing is the process by which an
organization determines the specific type of work-family policy or practice that might best
assuage the pressures it faces. This box is primarily informed by the agency perspective, which
attributes choice to organizations. This perspective suggests that there are many potential ways
for organizations to react to similar institutional pressures, and various organizational factors will
contribute to precisely how an organization will react to them. The first critical aspect of this
choice is the type of policy or practice to consider offering. In this model, I use one existing
typology of work-family practices developed by Kossek & Michel (2011), but any typology of
work-family policies and practices, including informal ones, may fit within the typing box. The
key is that the typology includes policies that address the pressures specific to the organization.

137

Though the possible combinations of pressures and policies are too numerous to discuss in depth,
I provide several examples.
The simplest such example is an organization that faces relatively little labor market
demand but high public policy demand. Such an organization might be in a context in which the
government requires organizations to provide many work-family practices, making it more
difficult to use them as a source of competitive advantage but also creating less labor market
pressure due to its needs being largely met by those practices. With the key pressure therefore
being public policy, an organization would choose practices that directly address the public
policy mandate. As another example, consider an organization experiencing little public policy
pressure, but high pressure from the labor market. The organization might then attempt to offer a
variety of policies to attract a diverse pool of talented applicants and better differentiate itself
from competitors.
A more complex example is an organization facing both pressures. In the best case
scenario, the talent it seeks to recruit away from its competitors will tend to place a high value on
certain work-family benefits that are required by law, e.g. sick leave. If the organization’s
competitors are not offering extensive sick leave policies, it can meet both pressures with a
generous sick leave policy that is more extensive or easier to use for the target applicant pool. If
the law requires a policy that is less directly related, for example paid maternity leave in a largely
male labor pool, the organization will likely have to offer multiple policies and practices.
A similarly complex but likely example is the organization actively seeking to
competitively recruit using work-family policies and practices. In order to provide a policy or
practice that can accomplish this goal, it must be desirable to the target labor pool. Thus, the
primary pressure comes from the labor market demands. In order for there to be room for

138

competitive advantage, similar policies must not be in place by competitors or mandated by
public policy (Piszczek & Berg, forthcoming). The policy type chosen will be the one that best
meets the target labor pool’s demands and differentiates the organization from its competitors.
The typing box and its ties to the environment box have close links to the existing
strategic human resource management literature. Chapter 3, grounded in this literature, provides
an example. In Chapter 3, I argue that there is a lack of public childcare in Germany, but that
childcare is desirable to working parents. Thus, I argue that organizations can leverage these
policies to reduce their organizational turnover and create strategic value. The results suggest
that organizations are adopting childcare practices only when turnover is already high, i.e. in
response primarily to labor market demand rather than to generate competitive advantage. But,
by showing a decrease in turnover following the adoption of a childcare practice within an
organization, the study demonstrates that these labor market demands are being met.
The interchangeable nature of the work-family policy and practices typology in this box
is echoed when using this framework expressly for research. Given the numerous content-based
classification systems, their usefulness in research will largely depend on the theoretical context
of the study. For example, a study grounded in boundary control theory may wish to differentiate
policies and practices by the degree employees tend to use them to either integrate or segment
work-family roles. In a study of telecommuting which places importance on both temporal and
physical flexibility, these categories might be combined. Researchers should make theoretically
grounded decisions of what typology to use if including this framework in a study.

139

Figure 4.1: An integrated model of the organizational work-family response
Environment

High

Low

Institutional
Pressure

Policy Typing





Low

High

Cost-Effectiveness of
Work-Family Practices

Low

Cost-Effectiveness of
Work-Family Practices
High

Strategic Value

Temporal Flexibility
Physical Flexibility
Workload Flexibility
Employment Breaks

Strategic
Adoption

Symbolic
Adoption

Compromised
Adoption

No Adoption

Strong

Weak

Mixed

None

Implementation

140

Many factors will determine the policy typing determined to best match the
organization's needs aside from institutional pressures. One such factor, implicit in the
consideration of the source of institutional pressure, is the target of the policy or practice.
Organizations often seek to attract diverse applicant pools which may have a variety of workfamily needs, and work-family policies and practices should thus offer a variety of practices to
satisfy various demographics in the labor market. On the other hand, if the pressure is from
public policy, the target employees are usually well-defined by the legislation. Another important
consideration for typing is the nature of industry and resulting organization of work in the
organization. For example, policies such as flexible shifts are not plausible for employees whose
work is not organized by shift. Similarly, healthcare organizations often require 24/7 staffing and
so flexible work hours may not be an option if it would result in understaffing in unpopular
hours. Knowledge-based work may be more amenable to telecommuting whereas other work
requires physical presence in the work-place.
These two examples of factors which help determine how well a policy or practice
matches an organization's needs (policy targets and organization of work) suggest contingencies
in the effectiveness of work-family policies and practices. When policies and practices that do
not match these potential contingencies are adopted, organizational and individual outcomes may
suffer. For example, if an organization decides to implement flexible shifts but very few of its
employees work on a shift system, the resulting impact on the organization may be negligible.
Many other factors such as organizational size and sector (i.e. public or private) likely contribute
to the match between a policy or practice and organization goals. As research continues to
explore cross-level relationships examining the boundaries of work-family policy effectiveness,
more of these factors will be identified. Moving through the framework, at this point an

141

organization has identified the degree to which it is experiencing pressure to adopt work-family
policies and practices, what plausible practices might meet that pressure, and who to target the
practices to in order to meet it.
The third box, strategic value, considers the actual net worth to the organization of the
chosen policy or practice. This incorporates the rational choice perspective discussed in Chapter
1 by considering whether the policy or practice chosen through policy typing has the potential to
help the organization's bottom line. As such, strategic value of a policy or practice is assessed by
whether the organization expects it to be cost-effective. Based on the institutional pressure
perspective, the importance of the perceived cost-effectiveness of a policy or practice is
intensified by the pressure an organization experiences to adopt it. Thus, in the strategic value
box, the expected cost-effectiveness for work-family practices and policies appears twice: once
under organizations experiencing high institutional pressure and once under those experiencing
low pressure.
This is similar to the institutional pressure perspective which creates a two by two matrix
of institutional pressure and expected technical outcomes (Goodstein, 1994), however I do not
adopt all its descriptions of organizational work-family response strategies, which do not
immediately link to the extent a specific policy or practice is implemented but rather broadly
describe the way policies and practices are offered holistically. The institutional pressure
framework used the terms acquiescence, avoidance, compromise, manipulation, and defiance to
describe the broad response patterns of organizations to institutional pressure to adopt workfamily practices.
Instead of using concepts from these frameworks, I use four categories which include
aspects of these holistic responses, but describe how a specific policy or practice is to be

142

adopted. These adoption styles are directly tied to the extent of a specific policy's or practice's
implementation, the subject of the fourth box. As such, I describe strategy and implementation
together. Implementation of a policy or practice has three key dimensions that will determine
how it affects employees: policy awareness, policy access, and policy extent. Each form of
adoption leads to a specific style of implementation along these three dimensions.
The first adoption style is strategic adoption. Strategic adoption occurs when the pressure
to adopt a particular practice or policy is high, and the organization expects such adoption to be
cost-effective. In this case, there is no immediate perceivable downside to implementing the
policy or practice and the organization should seek the strongest level of implementation
possible. Strong implementation is characterized by high awareness of a practice or policy
among employees, broad coverage of the policy granting many employees access to it, and high
practice quality in terms of its relative generosity to employees compared to those offered by
other organizations. An organization that is able to utilize strategic adoption can also expect the
most desirable results from such implementation for both employees and the organization. These
outcomes will be dependent on the policy (e.g., T. D. Allen, Johnson, Kiburz, & Shockley, 2013;
Kossek, Pichler, Bodner, & Hammer, 2011).
The second adoption style is symbolic adoption. When institutional pressure to adopt a
particular work-family policy or practice is high, organizations are more compelled to offer such
policies and practices and may find it difficult to offer nothing. However, if they expect these
policies and practices to be hurtful to the organization, they may opt to only partially implement
them or to implement them to as little an extent as possible. This can occur if the policies or
practices determined to best face the pressures are financially expensive or require additional
coordination or staffing that would outweigh the benefits. Organizations considering symbolic

143

adoption expect heavy fixed or recurring expenses with a policy or practice, extreme
coordination problems, or cultural pushback from such policies and will seek to minimize these
costs. Weak implementation may include symbolic or token offering of work-family policies or
practices, especially those required by law. Weakly implemented practices are characterized by
minimal employee awareness of their existence, narrow coverage of employees, difficulty of
access for those few that are covered, and minimal benefits. Organizations can have policies and
practices on paper but achieve weak implementation in a variety of ways. Supervisors, often
viewed as an extension of the organization by employees, may make policies difficult to access
(Casper, Fox, Sitzmann, & Landy, 2004; Eaton, 2003). Similarly, they may stigmatize
employees that are able to access work-family benefits as uncommitted to the organization
(Greenhaus & Kossek, 2014; Hewlett, Sherbin, & Forster, 2010).
The third category is compromised adoption. In this category, organizations expect
benefits of adopting work-family policies and practices, but such benefits may be limited by low
demand for them. Organizations in this category may be sector leaders, adopting minimal levels
of policy which ultimately create competitive pressure on other organizations to do the same.
However because an organization in this category faces little such pressure, it can expect
desirable results of implementing a policy less extensively than an organization facing high
institutional pressure. Thus, the level of implementation is mixed. Because the policy and
practice is expected to be beneficial to organizational outcomes, the key difference is likely to be
in the quality of the policy or practice rather than its ease of use or employee awareness of it.
Mixed implementation is thus characterized primarily by broad-reaching policies and practices
of limited generosity. Given the limited demand for such policies and practices, marginal returns
on their adoption are low. A rare exception may be the case in which an organization is the first

144

to adopt a work-family policy or practice that was not in high demand by the labor market but
subsequently creates that demand. Such cases, however, are few and far between and grow ever
more unlikely as work-family policies and practices become more common.
The final adoption form, no adoption, is unique and thus appears somewhat separately
from the other three in the model. It is unique in that it represents the absence of policies and
practices. When organizations face little or no pressure to adopt work-family policies or
practices, and they expect such policies and practices to be prohibitively costly either financially
or otherwise, they will offer no work-family policies and practices at all. This is partially
consistent with Kirchmeyer’s (1995) separation response category in that the organization
ignores the non-work needs of employees. This may be either due to ignorance resulting from
insufficient institutional pressure to bring such needs to the organization’s attention, or it may be
purposeful as part of a strategy to manage employee working time.
Chapter 2 provides an example of the latter. Organizations which have high after-hours
electronic communications expectations are able to compel employees to work during nonstandard hours outside of the workplace because there is little pressure on them to provide a
culture that values non-work time. These organizations may also consider the productivity costs
of having lower after-hours expectations not to be worth the increase in employee well-being
these lower expectations might provide. Thus organizations facing little-to-no institutional
pressure and that perceive work-family policies and practices as potentially having a net
undesirable effect on business outcomes will not offer any, and may even attempt to probe into
employees’ non-work time.
Overall this integrated framework answers the call to bridge the macro-micro divide
(Huselid & Becker, 2010) and adds organization-level theory to the work-family literature,

145

building on previous frameworks of the organizational work-family response in a way that is
compatible with current research. It links organization-level factors such as institutional pressure
and strategy to the actual adoption of a policy. It also explains differences in implementation of
similar policies across organizations, and why within organizations employees may have
different levels of access to work-family policies and practices. Next I discuss how this
framework might contribute to and be used in future studies.
4.5

Using the New Work-Family Response Framework
Overall, this updated framework provides more detail and a multi-stage understanding of

what organizational responses to employees’ work-family needs look like by explaining why
policies and practices are adopted and the degree to which they are implemented. It integrates
previous organizational frameworks and builds upon them to accommodate knowledge from
more current research. This updated framework yields a number of possible opportunities for
future research and practice.
The key contribution of this framework is that it links organization and contextual
variables to the adoption and implementation of policies and practices. In doing so, it also
connects organizational and contextual variables to individual and organizational outcomes
associated with those practices. Such associations are plentiful as the investigation of the
relationship between specific practices and their outcomes have been a primary focus of workfamily studies over the past twenty years. In this sense, the biggest contribution of the framework
is that it bridges the gap between employer and employee focused work-family research.
By bridging this gap, this framework helps researchers choose more theoretically relevant
outcomes, i.e. those that best address the pressures an organization is facing. By thinking about
practices and policies in terms of why they are adopted rather than their place in a typology of
146

practices, organizations and researchers can better measure their success by choosing outcomes
most relevant to organizational goals. Though work-family practices are understood to increase
employee attraction and retention, some organizations may be more focused on attraction and
others retention. This also highlights the mechanisms through which specific policies and
practices might lead to desirable organizational outcomes: by satisfying specific institutional
pressures. Though many studies now use multi-level modeling to examine potential
organizational or environmental influences, theoretical links from higher order constructs to the
employee level are not always entirely clear as research has focused more on whether workfamily policies work as intended rather than how they work (Odle-Dusseau et al., 2012). This
framework reintroduces institutional pressures as potential higher level constructs which can
influence the types of policies and practices implemented and the extent to which they're
implemented, as described above.
Similarly, in linking important environmental and organizational characteristics to the
adoption of policies and practices, the framework also helps identify boundary conditions around
the effectiveness of these practices and calls for more careful consideration of organizational
context in future research. Organizational and environmental contextual factors may help explain
the results of previous studies, such as the unexpected positive between-establishment
relationship between turnover and childcare policy prevalence in Chapter 3, or help guide the
selection of potential moderator variables which may impact a policy's effectiveness in future
studies. For example, a future study investigating the effectiveness of childcare policies may
wish to include a measure of experienced turnover, recruitment problems, or industry prevalence
of such policies, as the policy's effect may be stronger when turnover and recruitment are
problematic or when such policies are less common among competitors. Better identification of

147

these boundary conditions can help explain why policies and practices may be beneficial to one
organization but not another by putting a greater focus on the match between a policy or practice
and its ultimate goal in the organization rather than on how an average employee experiences it.
Emphasizing the match between policy or practice and the pressure it is meant to address
also helps practitioners and researchers define a successful implementation. It may help
practitioners decide which policies and practices are plausible in a real-world situation and how
to implement them. Practitioners can follow the figure from the top down to identify the
pressures their organization is facing, the expected results of a policy or practice within the
organization, and based on those factors, the degree to which the policy should be implemented.
They may even choose to implement multiple practices and policies, perhaps differentiating
between target employee groups, based on the pressures they face from different labor market
segments, whether a group of employees is core to the business, or whether the labor market is
shallow for certain skills. Once the appropriate institutional pressures, expected cost-benefit
ratio, and desired degree of enactment are identified, practitioners can much more easily ensure
that policies and practices are enacted consistently and synergistically with the organization's
goals at the forefront.
Another key contribution of this framework for future research its ability to account for
inconsistencies in work-family policies and practices both between and within organizations.
Above, I mention that environmental and organizational factors included in the framework may
explain inconsistent findings of relationships between specific work-family policies and practices
and individual and organizational outcomes. Additionally, because the framework can be applied
individually to multiple policies and practices, it can explain different levels of consistency in the
offering of a policy within an organization. For example, the offering of a policy to one group of

148

employees but not another may be because only the former group was a source of pressure. This
pressure may be a result of either the demographics of that employee segment or heavy
competition for it, or perhaps both. But without considering the factors described in this
framework, researchers are left to speculate on the source of these inconsistencies. This
framework explains variation in implementation, allowing multiple policies and practices to be
implemented differently even within a single organization. Strategic human resource
management studies can use this framework to better identify policy contingencies and
configurations (Delery & Doty, 1996) which may hold greater synergistic strategic value by
considering the specific pressures they address.
Finally, this framework facilitates comparative study of work-family policies and
practices from a different viewpoint. Researchers can use this framework to compare policies
and practices, or configurations of policies and practices, by how well they meet specific
institutional pressures and organizational goals rather than how they, for example, affect
individual role boundaries. These types of comparisons are more useful for organization-focused
research that seeks to better understand how work-family policies can be applied in a mutually
beneficial way that helps employers as well as employees. This also helps answer the recent call
of researchers to tie work-family policies more strongly to organizational goals (Kelly et al.,
2008; Kossek, Baltes, et al., 2011).
The framework also identifies theoretical gaps that could benefit from further
investigation. Future research aimed at the development of this framework should focus on
further explicating the typing process. The framework is purposefully broad so as to link broad
literatures, but I provide only a few examples of how experienced institutional pressures and
organizational variables such as work organization and institutional pressure source contribute to

149

the policy typing process. Other pressures may exist in addition to those described here. For
example, the need for legitimacy is an institutional pressure largely important for recruitment
(Collins, 2007). Additionally, there are many typologies of work-family policies and practices,
with no clear “best” typology. This complicates the typing process, as the dimensions used to
parse policies and practices can vary substantially across studies depending on their theoretical
grounding. As multi-level and organizational work-family studies become more feasible and
popular, these gaps in the framework will be easier to fill as researchers will better understand
the factors that go into the decision to implement a work-family policy or practice.
4.6

Limitations and Future Research
Though this updated framework adds much-needed detail to the organizational work-

family response framework, it has a few limitations. First and most importantly, this framework
is not exhaustive. As mentioned above, there is much that goes into the decision of whether or
not to adopt a particular work-family policy or practice, and all of those factors may not be
present in this framework. Though this framework incorporates what previous research suggests
are the most important factors (institutional pressures and cost-benefit expectations), other
factors may be important as well. Additionally, all institutional pressures and organizational
factors which impact the decisions of whether or not to adopt a policy or practice cannot be
accounted for in such a broad framework. In the text I describe key institutional pressures (labor
market demands and public policy) and organizational factors (organization of work and
experienced pressure sources) but other pressures and factors likely contribute to this decision as
well. For example, some research specifies the desire for legitimacy as an important institutional
pressure (DiMaggio & Powell, 1991) and previous institutional pressure research suggests that

150

public sector firms consider their duty to the public in their decision of whether to offer workfamily policies and practices (Goodstein, 1994).
Additionally, this framework is broad in its scope. Though this allows it to reach across
theoretical levels from environment to individual practice, it also means that there is less nuance
in the middle. More research is needed to better understand the typing process and how
organizations decide whether to adopt specific work-family policies and practices, if any, and
whom to apply them to. On a related note, this framework does not endorse a particular typology
of work-family policies or practices. Though this may contribute to the openness of the center
boxes, it also allows the framework to be compatible with any existing categorization or
theoretical breakdown of work-family policies and practices.
Future research should use this framework as a starting point to clearly and consistently
link organizational variables to individual and organizational outcomes through practices and
policies, and help explain why a particular policy or practice is implemented or not and to what
extent. Similarly, the framework should be used to help infer what the motivation behind a
particular work-family practice or policy (or lack thereof) might be, especially when this is not
clear from the population under study.
4.7

Final Conclusions
This dissertation consists of four chapters. Chapter 1 introduced the concept of the

organizational work-family response. Chapters 2 and 3 were empirical studies of two very
different forms of this response. These two studies showed the importance of the organization’s
method of responding to employee work-family needs but also demonstrated the need for a
broader framework for identifying and organizing these responses. This fourth and final chapter

151

uses these and other studies to update the work-family response framework and contextualize the
two individual empirical studies.
Overall, the organizational work-family response framework has not been consistent with
current research and lacks a theoretical perspective broad enough to capture the complexity of
organizational work-family policies and practices yet detailed enough to account for the
intricacies and variation in how and why they are adopted and implemented. This chapter
updated and reconceptualized the organizational work-family response as an integrated multistage framework with four primary components, creating a more in-depth map of how this
response is formed within organizations and what it might look like. This conceptualization
allows for clearer links from organizational goals to organizational and individual-level
outcomes by linking organization-level phenomenon such as institutional pressured more directly
to practices and policies which are better understood at the individual level.
Further, the integrated framework is more consistent with current research and trends in
the modern organization and distribution of work such as technology and globalization. This
updated framework also provides grounding contextualization for examining organizational
work-family policies and practices, facilitating comparisons within and between organizations. It
provides a starting point for researchers to examine not only the general type of work-family
response an organization presents but also its consistency. It provides a framework that grounds
organizational work-family practices in a context of other potential organizational choices. The
framework itself is grounded in the idea that organizations respond to the non-work needs of
employees, and it allows for comparisons of different potential ways an organization could
handle a work-family problem.

152

In discussing the benefits of this framework, I have also implicitly argued for a return to
research at a higher theoretical level. Though micro-level work-family research is thriving, and
meso-level research is growing, organizational work-family research has not received the same
attention. While from a practical standpoint organization-level research is difficult to conduct
due to the financial resources required and difficulty in obtaining quality data, it is necessary to
solve the puzzles of which work-family policies and practices work in what way, when they
work that way, and why they work that way. When organization-level factors are not focal to the
research question, or when organization-level data is too difficult to obtain, work-family
researchers can still contribute to our collective understanding of how organizations choose and
implement work-family policies and practices and how these decisions affect the impact these
policies and practices have on employees. At minimum, researchers should provide a discussion
of the study context so readers can identify potential institutional boundaries and contingencies
which may affect the generalization of the results and their implications for future study.
Regardless of whether it is incrementally through richer contextual descriptions in research or
more explicitly through organization-level study, this dissertation in its entirety shows that
consideration of the organizational level is important. Through the use of frameworks such as the
integrated, multi-stage framework presented in this chapter, researchers can address existing
puzzles of the organization's role in the work-family interface and usher in an era of crossdomain, multi-level research.

153

APPENDICES

154

APPENDIX A

Work-Family Technology Survey and Instructions

155

GENERAL WORK AND DEMOGRAPHIC QUESTIONS
1. What is your e-mail address?
(This information will be used only to send you the second wave of the survey and to match your
answers from the first and second waves. Your e-mail address will not be shared with anyone,
including your employer.)
2. What is your gender?
3. What is your age?
4. What is your marital status?
5. On average, how many hours do you work per week as standard working hours?
6. On average, how many hours do you work per week outside of your standard working hours?
7. How many children do you have under the age of 18?
8. Is your current job in the human resources profession?
9. What industry do you work in? Select the closest match.
Agriculture, Forestry, and Fishing
Mining
Construction
Manufacturing
Transportation, Communications, Electric, Gas, and Sanitary Services
Wholesale Trade
Retail Trade
Finance, Insurance, and Real Estate
Services
Public Administration
WORK METHODS AUTONOMY
Response Options: 1-5 from “strongly disagree” to “strongly agree”
10. The job allows me to make decisions about what methods I use to complete my work.
11. The job gives me considerable opportunity for independence and freedom in how I do the
work.

156

12. The job allows me to decide on my own how to go about doing my work.
EMOTIONAL EXHAUSTION
Response options: every day, a few times a week, once a week, a few times a month, once a
month or less, a few times a year or less, never
13. How often do you feel emotionally drained from your work?
14. How often do you feel burned out by your work?
15. How often do you feel used up at the end of the workday?
TURNOVER INTENTIONS
Response options: Response Options: 1-5 from “strongly disagree” to “strongly agree”
16. You are seriously considering quitting your company for another employer.
17. During the next 12 months, you will probably look for a new job outside your current
employer.
ROLE IDENTITY
Response Options: 1-7 from “strongly disagree” to “strongly agree”
WORK ROLE IDENTITY
18. People see me as highly focused on my work.
19. I invest a large part of myself in my work.
FAMILY ROLE IDENTITY
20. People see me as highly focused on my family.
21. I invest a large part of myself in my family life.
personal life activities throughout the day.
WORK-TO-FAMILY INTEGRATION PREFERENCE
Response Options: 1-7 from “strongly disagree” to “strongly agree”
22. I don't like to have to think about work while I'm at home.
23. I prefer to keep work life at work.
157

24. I don't like work issues creeping into my home life.
25. I like to be able to leave work behind when I go home.
FAMILY-TO-WORK INTEGRATION PREFERENCE
Response Options: 1-7 from “strongly disagree” to “strongly agree”
26. I don't like to have to think about home while I'm at work.
27. I prefer to keep my home life at home.
28. I don't like home issues creeping into my work life.
29. I like to be able to leave home behind when I go home.
WORK-FAMILY CONFLICT
Response Options: 1-5 from “strongly disagree” to “strongly agree”
30. The demands of your work interfere with your family or personal time.
31. The amount of time your job takes up makes it difficult to fulfill your family or personal
responsibilities.
32. Things you want to do at home do not get done because of the demands your job puts on you
33. Your job produces strain that makes it difficult to fulfill your family or personal duties.
34. Due to your work-related duties, you have to make changes to your plans for family or
personal activities.
AFTER-HOURS ELECTRONIC COMMUNICATIONS EXPECTATIONS
Response options: 1-5 from “not at all true” to “completely true”
Instructions: To what extent are the following statements true of you and your situation?
35. My organization expects me to respond to after-hours electronic work communications
immediately.
36. My organization expects me to be available for the organization to contact me in off hours.
37. My organization expects me to watch for incoming electronic communications from work
after-hours.

158

38. My organization expects me to be reachable through electronic communication when I go on
vacation.
39. My organization expects me to check for electronic communications from work when I am
on vacation.
40. When I'm given work that I need to finish at home, my organization expects me to let my
boss know via electronic communication as soon as it's finished.
41. If I have important information about work after hours, my organization expects me to
electronically communicate it right away.
TECHNOLOGY USE
Response options: 1-5 scale from “not at all” to “to a great extent”
Note: Only starred items retained for full sample.
USE FOR WORK WHILE AT HOME
Instructions: For the following questions, consider your usage of technology such as e-mail, cell
phones, voice mails, and instant messaging over a typical month.
In a typical month, to what extent do you...
*42. ...use technology to keep up with work matters while at home?
43. ...use technology to communicate with your workplace outside of normal work hours?
*44. ...use technology to perform work tasks while at home?
45. ...respond to someone from work using technology to communicate with you when you're at
home?
*46. ...use technology to stay connected to your work no matter where you are?
47. ...use technology to stay connected to your work no matter what time it is?
*48. ...use technology to access information pertaining to work tasks while at home?
49. ...experience interruptions at home from work-related communications?
50. ...provide status reports after working hours via technology?
51. ...check for technology-based communications as soon as you got up in the morning?
159

52. ...use technology to contact others in the organization to work on problems you learned about
after typical working hours?
53. ...actively check for messages using technology after typical working hours?
54. ...use technology to immediately respond to work-related communications after typical
working hours?
USE FOR HOME WHILE AT WORK
55. ...use technology to keep up with family matters while at work?
*56. ...use technology to communicate with your family or friends during normal work hours?
57. ...use technology to perform personal tasks while at work?
*58. ...respond to a family member using technology to communicate with you when you're
working?
59. ...use technology to stay connected to your family no matter where you are?
60. ...use technology to stay connected to your family no matter what time it is?
*61. ...use technology to access information pertaining to your family while working?
62. ...experience interruptions while working from family-related communications?
*63. ...check in with family during working hours via technology?
64. ...check for technology-based communications from family as soon as you get to work?
65. ...use technology to address family-related problems during the workday?
66. ...actively check for messages from family using technology after typical working hours?
67. ...use technology to immediately respond to family-related communications during typical
working hours?

160

APPENDIX B

Supplemental Analyses of German Childcare Practices

161

Table B.1: Within-establishment poisson regression models with establishment fixed effects
Total Turnover
Male Turnover
Female Turnover
Β
SE
IRR
Β
SE
IRR
Β
SE
IRR
Proportion of
employees
0.172
0.439 1.187
-0.426
0.481 0.653
0.659
0.533 1.934
female
Gross
0.000
0.000 1.000
0.000
0.000 1.000
0.000
0.000 1.000
monthly pay
Overtime
0.032
0.061 1.032
0.037
0.065 1.037
0.024
0.082 1.024
Collective
**
**
0.081 0.806 -0.233
0.080 0.792 -0.187 0.109 0.830
-0.216
agreement
Works
0.230
0.149 1.259
0.312
0.197 1.366
0.167
0.155 1.182
council
No equal
*
opportunity
-0.085
0.069 0.919 -0.135
0.067 0.874 -0.037 0.090 0.964
policies
*
**
Profit
0.020 0.960 -0.063
0.022 0.939 -0.021 0.026 0.979
-0.040
Business
0.037
0.044 1.038
-0.011
0.050 0.989
0.069
0.051 1.072
volume
Average
weekly work
0.000
0.002 1.000
0.002
0.002 1.003 -0.004 0.003 0.996
hours
Childcare
*
*
*
0.089 0.808 -0.189
0.078 0.828 -0.246
0.124 0.782
-0.214
practices
Parental
leave contact
0.003
0.088 1.004
-0.072
0.061 0.930
0.072
0.120 1.075
practices
Female
promotion
0.039
0.070 1.040
0.055
0.081 1.057
0.008
0.078 1.008
practices
*
Year: 2004
-0.032
0.034 0.968
0.039
0.036 1.040 -0.109
0.044 0.896
Year: 2008
N:
Observations
N: Groups
*

-0.154

**

*

0.063

0.857

-0.093

0.051

0.912

-0.230

*

0.101

17,369

14,022

11,608

7,274

5,876

4,914

Notes: p≤0.05; p≤0.01
Total number of employees set as exposure variable

162

0.795

Table B.2: Between-establishment poisson regression models
Total Turnover
Male Turnover
B
SE
IRR
B
SE
IRR
Proportion
of
**
-0.007
0.009 0.993 -1.777
0.013 0.169
employees
female
Gross
**
**
monthly pay 0.000
0.000 1.000 0.000
0.000 1.000
(10,000)
**
**
Overtime
0.006 1.342 0.302
0.008 1.353
0.294
Collective
**
**
0.006 1.021 0.036
0.008 1.037
0.020
agreement
Works
**
**
0.005 1.970 0.764
0.007 2.148
0.678
council
No equal
**
**
opportunity -0.143
0.007 0.867 -0.122
0.010 0.885
policies
**
**
Profit
0.002 0.883 -0.134
0.003 0.875
-0.124
Business
**
**
0.003 0.936 -0.027
0.004 0.974
-0.066
volume
Average
**
**
weekly
0.000 0.997 -0.002
0.000 0.998
-0.003
work hours
Childcare
**
**
0.006 1.210 0.224
0.009 1.251
0.191
practices
Parental
leave
**
**
0.007 0.820 -0.194
0.009 0.824
-0.198
contact
practices
Female
**
**
promotion
0.007 0.701 -0.347
0.009 0.707
-0.356
practices
**
**
Constant
0.046 0.662 -0.523
0.018 0.593
-0.413
N
40,017
39,724
*

**

Female Turnover
B
SE
IRR
1.971

0.000
0.282

**

0.014

7.177

**

0.000

1.000

**

0.008

1.325

**

0.008

0.925

**

0.008

1.666

**

0.011

0.815

**

0.003

0.905

**

0.004

0.951

**

0.000

0.998

**

0.009

1.291

**

0.010

0.813

**

0.010

0.738

**

0.077
39,725

0.123

-0.077
0.510

-0.205
-0.100
-0.050
-0.002
0.255

-0.207

-0.304
-2.094

Notes: p≤0.05; p≤0.01
Total number of employees set as exposure variable.
State and year controls not shown in table but were included in all analyses.

163

Within-Establishment

Table B.3: Within- and between-establishment combined poisson regression model
Female Turnover
Male Turnover
Variable
IRR
IRR
Β
SE
Β
SE
Proportion of employees
**
**
-0.385
0.895
female
0.083
0.680
0.082 2.448
Gross monthly pay (10,000) 0.000
0.000
1.000
0.000
0.000 1.000
**
Overtime
0.052
0.018
1.054
0.032
0.018 1.032
**
**
Collective agreement
-0.224
-0.190
0.020
0.799
0.023 0.827
Works council
No equal opportunity
policies
Profit
Business volume
Average weekly work hours
Childcare practices
Parental leave contact
Female promotion practices
Proportion of employees
female
Gross monthly pay (10,000)

Between-Establishment

Overtime
Collective agreement
Works council
No equal opportunity
policies
Profit
Business volume
Average weekly work hours
Childcare practices
Parental leave contact
Female promotion practices
Constant
N: Observations
N: Groups
*

**

0.317

**

**

0.035

1.372

0.124

0.015

0.882

-0.026

0.005

0.945

-0.015

0.013

0.949

0.059

0.000

1.002

-0.004

0.017

0.814

-0.270

0.015

0.929

0.055

0.016

1.078

0.023

0.096

0.181

1.549

0.000

1.000

0.000

-0.010
**
0.199

0.034

0.990

0.160

0.032

1.220

0.129

0.044

1.138

0.026
**
0.151

0.045

1.070

-0.121

0.011

0.866

-0.174

0.020

0.959

-0.178

0.001

0.996

0.002

0.058

1.554

0.419

0.048

0.906

-0.079

0.057

0.806

-0.233

-0.126
-0.057
-0.052
0.002

-0.074

**
**

**

**

**

0.067
-0.144
-0.042
-0.004
0.441

**

**

-1.711
0.000

**

**

-0.206
0.075

**

**
*
**

**

-0.098
-0.216
-1.283

*
**
**

-0.221 0.680
35,781
22,798

-3.02

**

**
**
**

**

**
*
**

*
**
**

*
**
**

**

0.032

1.132

0.016

0.975

0.006

0.986

0.011

1.061

0.001

0.996

0.019

0.763

0.015

1.056

0.017

1.024

0.097

4.708

0.000

1.000

0.037

1.174

0.037

1.027

0.044

1.163

0.049

0.886

0.013

0.841

0.020

0.837

0.001

1.002

0.063

1.520

0.052

0.924

0.062

0.792

0.080
35,783
22,799

2.448

Notes: p≤0.05; p≤0.01
Total turnover model does not converge.
Total number of employees set as exposure variable.
State, industry, and year controls not shown in table but were included in all analyses.
164

REFERENCES

165

REFERENCES

Abrahamson, P. (1999). The Male Bread-winner Model under Change: The Case of Germany
towards the 21st Century. Denmark: Roskilde University.
Allen, T. D. (2001). Family-Supportive Work Environments: The Role of Organizational
Perceptions. Journal of Vocational Behavior, 58(3), 414–435. doi:10.1006/jvbe.2000.1774
Allen, T. D., Cho, E., & Meier, L. L. (2014). Work–Family Boundary Dynamics. Annual Review
of Organizational Psychology and Organizational Behavior, 1(1), 99–121.
doi:10.1146/annurev-orgpsych-031413-091330
Allen, T. D., Johnson, R. C., Kiburz, K. M., & Shockley, K. M. (2013). Work-Family Conflict
and Flexible Work Arrangements: Deconstructing Flexibility. Personnel Psychology, 66(2),
345–376. doi:10.1111/peps.12012
Allen, V., & van de Vliert, E. (1984). A Role Theoretical Perspective on Transitional Processes.
In V. Allen & E. van de Vliert (Eds.), Role Transitions: Explorations and Explanations (pp.
3–18). New York, NY: Plenum Press.
Ammons, S. K. (2013). Work-family boundary strategies: Stability and alignment between
preferred and enacted boundaries. Journal of Vocational Behavior, 82(1), 49–58.
doi:10.1016/j.jvb.2012.11.002
Ashforth, B., Kreiner, G. E., & Fugate, M. (2000). All in a day’s work: Boundaries and micro
role transitions. Academy of Management Review, 25(3), 472–491.
Bailey, D. E., Leonardi, P. M., & Chong, J. (2010). Minding the Gaps: Understanding
Technology Interdependence and Coordination in Knowledge Work. Organization Science,
21(3), 713–730. doi:10.1287/orsc.1090.0473
Bakker, A. B., Demerouti, E., & Dollard, M. F. (2008). How job demands affect partners’
experience of exhaustion: integrating work-family conflict and crossover theory. The
Journal of Applied Psychology, 93(4), 901–11. doi:10.1037/0021-9010.93.4.901
Baltes, B., Briggs, T., Huff, J., Wright, J., & Neuman, G. (1999). Flexible and compressed
workweek schedules: A meta-analysis of their effects on work-related criteria. Journal of
Applied Psychology, 84(4), 496–513.
Bardoel, E., Tharenou, P., & Moss, S. (1999). Organizational Predictors of Work-Family
Practices. Asia Pacific Journal of Human Resources, 36(3), 31–49.
doi:10.1177/103841119903600304

166

Barley, S. R., Meyerson, D. E., & Grodal, S. (2011). E-mail as a Source and Symbol of Stress.
Organization Science, 22(4), 887–906. doi:10.1287/orsc.1100.0573
Baron, A. (2005). Build technology into strategy. Strategic Communication Management, 9(4),
13.
Baron, R., & Kenny, D. (1986). The moderator–mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6), 1173–1182.
BASF. (2013). Promoting Work-Life-Balance. Retrieved from
http://www.basf.com/group/corporate/en/sustainability/employees/nursery_lukids
Batt, R., & Valcour, P. M. (2003). Human Resources Practices as Predictors of Work-Family
Outcomes and Employee Turnover. Industrial Relations, 42(2), 189–220.
Bayer. (2013). Jobs and career. Germany. What we offer. Work-life balance. Retrieved from
http://www.mybayerjob.de/en/whybayer/offering/flexibleworktime
BBC. (2011). Volkswagen turns off Blackberry email after work hours. BBC News.
Beauregard, T., & Henry, L. (2009). Making the link between work-life balance practices and
organizational performance. Human Resource Management Review, 19(1), 9–22.
Becker, B. E., & Huselid, M. A. (1998). High performance work systems and firm performance:
A synthesis of research and manaerial implications. In G. R. Ferris (Ed.), Research in
Personnel and Human Resource Management (pp. 53–101). Greenwich, CT: JAI Press.
Becker, B. E., & Huselid, M. A. (2006). Strategic Human Resources Management: Where Do
We Go From Here? Journal of Management, 32(6), 898–925.
doi:10.1177/0149206306293668
Bellotti, V., Ducheneaut, N., Howard, M., & Smith, I. (2003). Taking email to task: the design
and evaluation of a task management centered email tool. In Proceedings of the SIGCHI
Conference (pp. 345–352). Ft. Lauderdale, FL.
Bellotti, V., Ducheneaut, N., Howard, M., Smith, I., & Grinter, R. E. (2005). Quality versus
quantity: E-mail-centric task management and its relation with overload. Human-Computer
Interaction, 20, 89–138.
Bennhold, K. (2010, January 17). In Germany, a Tradition Falls, and Women Rise. The New
York Times. New York, NY.
Berg, P., Kalleberg, A. L., & Appelbaum, E. (2003). Balancing work and family: The role of
high-commitment environments. Industrial Relations, 42(2), 168–188.

167

Berg, P., Kossek, E. E., Baird, M., & Block, R. N. (2013). Collective Barganing and Public
Policy: Pathways to Work-Family Policy Adoption in Australia and the United States.
European Management Journal, 31(5), 495–504.
Berg, P., Kossek, E. E., Misra, K., & Belman, D. (2014). Do Unions Matter for Work-Life
Flexibility Policy Access and Use? Industrial & Labor Relations Review.
Berg, P., & Piszczek, M. M. (2014). The limits of equality bargaining in the USA. Journal of
Industrial Relations. doi:10.1177/0022185613517469
Berk, R. A., & Beck, S. F. (1979). Labor and Leisure at Home. Beverly Hills, CA: SAGE
Publications.
Bianchi, S. M., Sayer, L. C., Milkie, M. A., & Robinson, J. P. (2012). Housework: Who Did,
Does or Will Do It, and How Much Does It Matter? Social Forces, 91(1), 55–63.
doi:10.1093/sf/sos120
Blackler, F. (1995). Knowledge, knowledge work and organizations: an overview and
interpretation. Organization Studies, 16(6), 1021–1046.
Bloom, N., Kretschmer, T., & Van Reenen, J. (2010). Are family friendly workplace practices a
valuable firm resource? Strategic Management Journal, 32(June), 343–367.
doi:10.1002/smj
Boettcher, A., & Krieger, S. (2008). Parental allowance - a first look back. Destatis: Statistiches
Bundesamt. Retrieved from
https://www.destatis.de/EN/Publications/STATmagazin/SocialBenefits/2008_10/2008_10P
arentalAllowance.html
Bollen, K. A., & Stine, R. (1990). Direct and indirect effects: Classical and bootstrap estimates
of variability. Sociological Methodology, 20, 115–140.
Boroff, K., & Lewin, D. (1997). Loyalty, voice and intent to exit a union firm: a conceptual and
empirical analysis. Industrial & Labor Relations Review, 51(1), 50–63.
Boswell, W. R., & Olson-Buchanan, J. B. (2007). The Use of Communication Technologies
After Hours: The Role of Work Attitudes and Work-Life Conflict. Journal of Management,
33(4), 592–610. doi:10.1177/0149206307302552
Bulger, C. A., Matthews, R. A., & Hoffman, M. E. (2007). Work and personal life boundary
management: boundary strength, work/personal life balance, and the segmentationintegration continuum. Journal of Occupational Health Psychology, 12(4), 365–75.
doi:10.1037/1076-8998.12.4.365
Bundesministerium fuer Familie, Senioren, Frauen, und J. (2007). Familien brauchen
Rahmenbedingungen: Deutschland wird familienfreundlicher. Berlin, Germany.
168

Bundesministerium fuer Familie, Senioren, Frauen, und J. (2010). Familien Report 2010:
Leistungen, Wirkungen, Trends. Berlin, Germany.
Bundesministerium fuer Familie, Senioren, Frauen, und J. (2013a). Das Elterngeld. Berlin,
Germany.
Bundesministerium fuer Familie, Senioren, Frauen, und J. (2013b). Das Kindergeld. Berlin,
Germany.
Bundesministerium fuer Familie, Senioren, Frauen, und J. (2013c). Kinderbetreuung:
Betreuungsgeld. Retrieved from http://www.familienwegweiser.de/wegweiser/stichwortverzeichnis,did=199554.html
Burke, M. E. (2004). Generational Differences. Alexandria, VA.
Butts, M. M., Casper, W. J., & Yang, T. S. (2013). How important are work-family support
policies? A meta-analytic investigation of their effects on employee outcomes. Journal of
Applied Psychology, 98(1), 1–25. doi:10.1037/a0030389
Callaghan, G., & Thompson, P. (2001). Edwards Revisited: Technical Control and Call Centres.
Economic and Industrial Democracy, 22(1), 13–37. doi:10.1177/0143831X01221002
Carless, S., & Wintle, J. (2007). Applicant Attraction: The role of recruiter function, work–life
balance policies and career salience. International Journal of Selection and Assessment,
15(4), 394.
Carlson, K. D., & Wu, J. (2012). The Illusion of Statistical Control: Control Variable Practice in
Management Research. Organizational Research Methods, 15(3), 413–435.
doi:10.1177/1094428111428817
Casper, W. J., & Buffardi, L. C. (2004). Work-life benefits and job pursuit intentions: The role of
anticipated organizational support. Journal of Vocational Behavior, 65(3), 391–410.
doi:10.1016/j.jvb.2003.09.003
Casper, W. J., De Hauw, S., & Wayne, J. H. (2013). Concepts and measures in the work-family
interface: implications for work-family integration. In D. A. Major & R. J. Burke (Eds.),
Handbook of Work-Life Integration Among Professionals: Challenges and Opportunities.
Northampton, MA: Edward Elgar Publishing.
Casper, W. J., Fox, K. E., Sitzmann, T. M., & Landy, A. L. (2004). Supervisor referrals to workfamily programs. Journal of Occupational Health Psychology, 9(2), 136–51.
doi:10.1037/1076-8998.9.2.136
Casper, W. J., & Harris, C. M. (2008). Work-life benefits and organizational attachment: Selfinterest utility and signaling theory models. Journal of Vocational Behavior, 72(1), 95–109.
doi:10.1016/j.jvb.2007.10.015
169

Chakrabarti, M. (2011). Work-family boundary management strategies: examining outcomes,
and the role of fit. Wayne State University.
Chen, G., Kanfer, R., DeShon, R. P., Mathieu, J. E., & Kozlowski, S. W. J. (2009). The
motivating potential of teams: Test and extension of cross-level model of motivation in
teams. Organizational Behavior and Human Decision Processes, 110(1), 45–55.
doi:10.1016/j.obhdp.2009.06.006
Chen, Z., Powell, G. N., & Greenhaus, J. H. (2009). Work-to-family conflict, positive spillover,
and boundary management: a person-environment fit approach. Journal of Vocational
Behavior, 74(1), 82–93. doi:10.1016/j.jvb.2008.10.009
Chesley, N. (2005). Blurring boundaries? Linking technology use, spillover, individual distress,
and family satisfaction. Journal of Marriage and Family, 67(December), 1237–1248.
Chesley, N. (2006). Families in a high-tech age technology usage patterns, work and family
correlates, and gender. Journal of Family Issues, 27(5), 587–608.
Clark, R., Nye, F., & Gecas, V. (1978). Husbands’ work involvement and marital role
performance. Journal of Marriage and the Family, 40(1), 9–21.
Clark, S. C. (2000). Work/Family Border Theory: A New Theory of Work/Family Balance.
Human Relations, 53(6), 747–770. doi:10.1177/0018726700536001
Cleveland, J. N., & McCarthy, J. (2013). Work and family issues in a multi-generational context.
In D. A. Major & R. J. Burke (Eds.), Handbook of Work-Life Integration Among
Professionals: Challenges and Opportunities. Northampton, MA: Edward Elgar Publishing.
Cohen, A. (1997). Personal and Organizational Responses to Work-Nonwork Interface as
Related to Organizational Commitment. Journal of Applied Social Psychology, 27(12),
1085–1114.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Interactions Among Continuous
Variables. In Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences
(3rd ed., pp. 255–301). Hillsdale, NJ: Lawrence Erlbaum Associates.
Collins, C. J. (2007). The interactive effects of recruitment practices and product awareness on
job seekers’ employer knowledge and application behaviors. The Journal of Applied
Psychology, 92(1), 180–90. doi:10.1037/0021-9010.92.1.180
Combs, J., Liu, Y., Hall, A., & Ketchen, D. (2006). How much do high performance work
practices matter? A meta-analysis of their effects on organizational performance. Personnel
Psychology, 59, 501–528.

170

Cook, C., Heath, F., & Thompson, R. L. (2000). A Meta-Analysis of Response Rates in Web- or
Internet-Based Surveys. Educational and Psychological Measurement, 60(6), 821–836.
doi:10.1177/00131640021970934
Cook, T. D., & Campbell, D. T. (1979). Quasi-Experimentation: Design and Analysis for Field
Settings. Chicago, IL: Rand McNally.
Costello, A., & Osborne, J. (2005). Best practices in exploratory factor analysis: four
recommendations for getting the most from your analysis. Practical Assessment, Research
& Evaluation, 10(7), 1–9.
Dalton, D., & Mesch, D. (1990). The impact of flexible scheduling on employee attendance and
turnover. Administrative Science Quarterly, 35(2), 370–387.
Davis, A. (2013). These Stores Refuse To Open On Thanksgiving. CBS Minnesota. Retrieved
from http://minnesota.cbslocal.com/2013/11/26/these-stores-refuse-to-open-onthanksgiving/
Dawley, D. D., & Anthony, W. P. (2003). User Perceptions of E-Mail at Work. Journal of
Business and Technical Communication, 17(2), 170–200. doi:10.1177/1050651902250947
Dawson, J. F., & Richter, A. W. (2006). Probing three-way interactions in moderated multiple
regression: development and application of a slope difference test. The Journal of Applied
Psychology, 91(4), 917–26. doi:10.1037/0021-9010.91.4.917
Day, A., Scott, N., & Kelloway, E. (2010). Information and communication technology:
Implications for job stress and employee well-being. In New Developments in Theoretical
and Conceptual Approaches to Job Stress: Research in Occupational Stress and Well
Being, Volume 8 (pp. 317–350).
Dedrick, J., Gurbaxani, V., & Kraemer, K. L. (2003). Information Technology and Economic
Performance : A Critical Review of the Empirical Evidence, 35(1), 1–28.
Delery, J., & Doty, D. (1996). Modes of theorizing in strategic human resource management:
Tests of universalistic, contingency, and configurational performance predictions. Academy
of Management Journal, 39(4), 802–835.
Demerouti, E., Bakker, A. B., & Bulters, A. J. (2004). The loss spiral of work pressure, work–
home interference and exhaustion: Reciprocal relations in a three-wave study. Journal of
Vocational Behavior, 64(1), 131–149. doi:10.1016/S0001-8791(03)00030-7
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demandsresources model of burnout. Journal of Applied Psychology, 86(3), 499–512.
Den Dulk, L., Groeneveld, S., Ollier-Malaterre, A., & Valcour, M. (2013). National context in
work-life research: A multi-level cross-national analysis of the adoption of workplace work171

life arrangements in Europe. European Management Journal, 31(5), 478–494.
doi:10.1016/j.emj.2013.04.010
Derks, D., & Bakker, A. (2010). The impact of e-mail communication on organizational life.
Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 4(1).
Derks, D., & Bakker, A. B. (2012). Smartphone use, work-home interference, and burnout: A
diary study on the role of recovery. Applied Psychology: An International Review, 1–30.
doi:10.1111/j.1464-0597.2012.00530.x
Desrochers, S. (2005). Preliminary Validation of the Work-Family Integration-Blurring Scale.
Journal of Family Issues, 26(4), 442–466. doi:10.1177/0192513X04272438
Desrochers, S., & Sargent, L. (2004). Boundary/Border Theory and Work-Family Integration.
Organization Management Journal, 1(1), 40–48.
Diaz, I., Chiaburu, D. S., Zimmerman, R. D., & Boswell, W. R. (2012). Communication
technology: Pros and cons of constant connection to work. Journal of Vocational Behavior,
80(2), 500–508. doi:10.1016/j.jvb.2011.08.007
DiMaggio, P. J., & Powell, W. W. (1991). The iron cage revisited: institutional isomorphism and
collective rationality in organization fields. In P. J. Powell & W. W. DiMaggio (Eds.), The
new institutionalism in organizational analysis (pp. 41–63). Chicago, IL: University of
Chicago Press.
Duxbury, L., Towers, I., Higgins, C., & Thomas, A. (2006). From 9 to 5 to 24 and 7: How
Technology Redefined the Work Day. In W. Law (Ed.), Information Resources
Management: Global Challenges (pp. 305–332). Hershey, PA: Idea Group Publishing.
Eaton, S. (2003). If you can use them: Flexibility policies, organizational commitment, and
perceived performance. Industrial Relations, 42(2), 145–168.
Edwards, J., & Rothbard, N. (1999). Work and family stress and well-being: An examination of
person-environment fit in the work and family domains. Organizational Behavior and
Human Decision Processes, 77(2), 85–129.
Edwards, R. (1979). Contested Terrain: The Transformation of the Workplace in the Twentieth
Century. London: Heinemann.
Edwards, R. (1984). Forms of Control in the Labor Process: An Historical Analysis. In
Organization and Bureaucracy (pp. 109–142). Philadelphia, PA: Temple University Press.
Evans, W. R. (2005). High-Performance Work Systems and Organizational Performance: The
Mediating Role of Internal Social Structure. Journal of Management, 31(5), 758–775.
doi:10.1177/0149206305279370

172

Fabrigar, L., Wegener, D., MacCallum, R., & Strahan, E. (1999). Evaluating the use of
exploratory factor analysis in psychological research. Psychological Methods, 4(3), 272–
299.
Fender, C. M. (2010). Electronic Tethering: Perpetual Wireless Connectivity to the
Organization. Drexel University.
Fenner, G. H., & Renn, R. W. (2004). Technology-assisted supplemental work: Construct
definition and a research framework. Human Resource Management, 43(2-3), 179–200.
doi:10.1002/hrm.20014
Fenner, G. H., & Renn, R. W. (2010). Technology-assisted supplemental work and work-tofamily conflict: The role of instrumentality beliefs, organizational expectations and time
management. Human Relations, 63(1), 63–82. doi:10.1177/0018726709351064
Fernandez-Alles, M., & Ramos-Rodríguez, A. (2009). Intellectual structure of human resources
management research: A bibliometric analysis of the journal human resource management,
1985–2005. Journal of the American Society for Information Science and Technology,
60(1), 161–175. doi:10.1002/asi
Fligstein, N., & Sharone, O. (2002). Work in the Postindustrial Economy in California. In R.
Milkman (Ed.), The State of California Labor (pp. 67–96). Berkeley: University of
California Press.
Frone, M. R., Russell, M., & Cooper, M. L. (1992). Prevalence of work-family conflict : Are
work and family boundaries asymmetrically permeable ?, 13(August 1991).
Frye, N., & Breaugh, J. (2004). Family-friendly policies, supervisor support, work–family
conflict, family–work conflict, and satisfaction: A test of a conceptual model. Journal of
Business and Psychology, 19(2), 197–221.
Galinsky, E., Aumann, K., & Bond, J. (2009). Times are changing: Gender and generation at
work and at home. New York, NY.
Galinsky, E., Sakai, K., Eby, S., Bond, J., & Wigton, T. (2010). Employer-Provided Workplace
Flexibility. In K. Christensen & B. Schneider (Eds.), Workplace Flexibility: Realigning 20th
Century Jobs for a 21st Century Workforce. Ithaca, NY: Cornell University Press.
Gittell, J. H., Seidner, R., & Wimbush, J. (2010). A Relational Model of How High-Performance
Work Systems Work. Organization Science, 21(2), 490–506. doi:10.1287/orsc.1090.0446
Glass, J., & Estes, S. (1997). The family responsive workplace. Annual Review of Sociology, 23,
289–313.

173

Glass, J., & Fujimoto, T. (1995). Employer Characteristics and the Provision of Family
Responsive Policies. Work and Occupations, 22(4), 380–411.
doi:10.1177/0730888495022004002
Golden, A. G., & Geisler, C. (2007). Work-life boundary management and the personal digital
assistant. Human Relations, 60(3), 519–551. doi:10.1177/0018726707076698
González, V. M., & Mark, G. (2004). “ Constant , Constant , Multi-tasking Craziness ”:
Managing Multiple Working Spheres, 6(1), 113–120.
Good Technology. (2012). Good Technology Survey Reveals Americans are Working More, but
on their Own Schedule. Retrieved from http://www1.good.com/about/pressreleases/161009045.html
Goodstein, J. D. (1994). Institutional Pressures and Strategic Responsiveness: Employer
Involvement in Work-Family Issues. Academy of Management Journal, 37(2), 350–382.
Goodstein, J. D. (1995). Employer involvement in eldercare: An organizational adaptation
perspective. Academy of Management Journal, 38(6), 1657–1671.
Gould, W. (2001). What is the between estimator? Stata.com. Retrieved June 03, 2014, from
http://www.stata.com/support/faqs/statistics/between-estimator/
Greenhaus, J. H., & Beutell, N. J. (1985). Sources of and Conflict Family Between Work.
Academy of Management Review, 10(1), 76–88.
Greenhaus, J. H., & Kossek, E. E. (2014). The Contemporary Career: A Work–Home
Perspective. Annual Review of Organizational Psychology and Organizational Behavior,
1(1), 361–388. doi:10.1146/annurev-orgpsych-031413-091324
Guthrie, J. (2001). High-involvement work practices, turnover, and productivity: Evidence from
New Zealand. Academy of Management Journal, 44(1), 180–190.
Hall, D. D. T., & Richter, J. (1988). Balancing work life and home life: what can organizations
do to help? Academy of Management Executive, 11(3), 213–223.
Hammer, L. B., Kossek, E. E., Yragui, N. L., Bodner, T. E., & Hanson, G. C. (2009).
Development and Validation of a Multidimensional Measure of Family Supportive
Supervisor Behaviors (FSSB). Journal of Management, 35(4), 837–856.
doi:10.1177/0149206308328510
Hank, K., & Kreyenfeld, M. (2003). A multilevel analysis of child care and women’s fertility
decisions in Western Germany. Journal of Marriage and Family, 65, 584–596.
doi:10.1111/j.1741-3737.2003.00584.x

174

Hawley, C. (2012, June). Letter from Berlin: Childcare Draft Law Fuels Feminism Debate in
Germany. Der Spiegel Online. Berlin.
Heidenreich, U. (2013). So kommen Eltern zu ihrem Recht. Sueddeutsche.de. Retrieved from
http://www.sueddeutsche.de/politik/anspruch-auf-kita-plaetze-so-kommen-eltern-zu-ihremrecht-1.1720915
Hewlett, S. A., Sherbin, L., & Forster, D. (2010, June). Off-ramps and on-ramps revisited.
Harvard Business Review.
Hill, E. J., Grzywacz, J. G., Allen, S., Blanchard, V. L., Matz-Costa, C., Shulkin, S., & PittCatsouphes, M. (2008). Defining and conceptualizing workplace flexibility. Community,
Work & Family, 11(2), 149–163. doi:10.1080/13668800802024678
Honeycutt, T. L., & Rosen, B. (1997). Family friendly human resource policies, salary levels,
and salient identity as predictors of organizational attraction. Journal of Vocational
Behavior, 50, 271–290. doi:10.1006/jvbe.1996.1554
Huselid, M. A. (1995). The impact of human resource management practices on turnover,
productivity, and corporate financial performance. Academy of Management Journal, 38(3),
635–672.
Huselid, M. A., & Becker, B. E. (1996). Methodological issues in cross-sectional and panel
estimates of the human resource-firm performance link. Industrial Relations, 35(3), 400.
Huselid, M. A., & Becker, B. E. (2010). Bridging Micro and Macro Domains: Workforce
Differentiation and Strategic Human Resource Management. Journal of Management,
37(2), 421–428. doi:10.1177/0149206310373400
Huselid, M. A., Jackson, S. E., & Schuler, R. S. (1997). Technical and strategic human resource
management effectiveness as determinants of firm performance. Academy of Management
Journal, 40(1), 171–188.
Ichniowski, C., & Shaw, K. (1999). The Effects of Human on Resource Economic Plants
Management Systems Performance: An of and Japanese International Comparison.
Management Science, 45(5), 704–721.
Ilies, R., Schwind, K. M., Wagner, D. T., Johnson, M. D., DeRue, D. S., & Ilgen, D. R. (2007).
When can employees have a family life? The effects of daily workload and affect on workfamily conflict and social behaviors at home. The Journal of Applied Psychology, 92(5),
1368–79. doi:10.1037/0021-9010.92.5.1368
Ingram, P., & Simons, T. (1995). Institutional and resource dependence determinants of
responsiveness to work-family issues. Academy of Management Journal, 38(5), 1466–1482.

175

Institute for Child and Family Policy. (2008). Germany. Retrieved from
http://childpolicyintl.org/countries/germany.html
Jacobs, J., & Green, K. (1998). Who are the overworked Americans? Review of Social Economy,
56(4), 442–459.
Jones, R., & Murrell, A. J. (2001). Signaling Positive Corporate Social Performance: An Event
Study of Family-Friendly Firms. Business & Society, 40(1), 59–78.
doi:10.1177/000765030104000105
Kakabadse, N. K., Porter, G., & Vance, D. (2009). The unbalanced high-tech life: are employers
liable? Strategic Change, 18(1-2), 1–13. doi:10.1002/jsc.837
Kanter, R. M. (1977). Work and family in the United States: A critical review and agenda for
research and policy. New York, NY: Russell Sage Foundation.
Karasek, R. (1979). Job demands, job decision latitude, and mental strain: Implications for job
redesign. Administrative Science Quarterlyy, 24(2), 285–308.
Katz, D., & Kahn, R. (1978). The social psychology of organizations (2nd ed.). New York, NY:
John Wiley & Sons.
Kaufman, B. (2010). SHRM Theory in the Post-Huselid Era: Why It Is Fundamentally
Misspecified. Industrial Relations, 49(2), 286–314.
Kaufman-Scarborough, C. (2006). Time Use and the Impact of Technology: Examining
workspaces in the home. Time & Society, 15(1), 57–80. doi:10.1177/0961463X06061782
Keeney, J. (2012). Job flexibility, work-nonwork interruptions, and implciations for worknonwork conflict. Michigan State University.
Kelly, E. L., Kossek, E. E., Hammer, L. B., Durham, M., Bray, J., Chermack, K., … Kaskubar,
D. (2008). Getting There from Here: Research on the Effects of Work-Family Initiatives on
Work-Family Conflict and Business Outcomes. Academy of Management Annals, 2(1),
305–349. doi:10.1080/19416520802211610
King, W. R., & He, J. (2006). A meta-analysis of the technology acceptance model. Information
& Management, 43(6), 740–755. doi:10.1016/j.im.2006.05.003
Kirchmeyer, C. (1995). Managing the Work-Nonwork Boundary: An Assessment of
Organizational Responses. Human Relations, 48(5), 515–536.
doi:10.1177/001872679504800504
Konrad, A. M., & Mangel, R. (2000). The Impact of Work-Life Programs on Firm Productivity.
Strategic Management Journal, 21(12), 1225–1237.

176

Kossek, E. E., Baltes, B. B., & Matthews, R. A. (2011). How work-family research can finally
have an impact in organizations. Industrial and Organizational Psychology, 4(3), 352–369.
doi:10.1111/j.1754-9434.2011.01353.x
Kossek, E. E., Dass, P., & DeMarr, B. (1994). The Dominant Logic of Employer-Sponsored
Work and Family Initiatives: Human Resource Managers’ Institutional Role. Human
Relations, 47(9), 1121–1149. doi:10.1177/001872679404700905
Kossek, E. E., & Friede, A. (2005). The business case: Managerial perspectives on work and the
family. In M. Pitt-catsouphes, E. E. Kossek, & S. Sweet (Eds.), The Handbook of WorkFamily: Multi-Disciplinary Perspectives, Methods, and Approaches (pp. 611–628).
Mahwah, NJ: Lawrence Erlbaum Associates.
Kossek, E. E., Hammer, L. B., Kelly, E. L., & Moen, P. (2014). Designing Work, Family &
Health Organizational Change Initiatives. Organizational Dynamics, 43, 53–63.
Kossek, E. E., & Lautsch, B. A. (2012). Work-family boundary management styles in
organizations: A cross-level model. Organizational Psychology Review, 2(2), 152–171.
doi:10.1177/2041386611436264
Kossek, E. E., Lautsch, B. A., & Eaton, S. C. (2006). Telecommuting, control, and boundary
management: Correlates of policy use and practice, job control, and work–family
effectiveness. Journal of Vocational Behavior, 68(2), 347–367.
doi:10.1016/j.jvb.2005.07.002
Kossek, E. E., Lautsch, B. A., & Eaton, S. C. (2009). “Good Teleworking”: Under What
Conditions Does Teleworking Enhance Employees’ Well-being? In Y. Amichai-Hamburger
(Ed.), Technology and Psychological Well-Being. Cambridge University Press.
Kossek, E. E., Lewis, S., & Hammer, L. B. (2010). Work-life initiatives and organizational
change: Overcoming mixed messages to move from the margin to the mainstream. Human
Relations, 63(1), 3–19. doi:10.1177/0018726709352385
Kossek, E. E., & Michel, J. S. (2011). Flexible Work Schedules. In S. Zedeck (Ed.), APA
Handbook of Industrial and Organizational Psychology Volume 1 (pp. 535–572).
Washington, DC: American Psychological Association.
Kossek, E. E., & Nichol, V. (1992). The effects of on-site child care on employee attitudes and
performance. Personnel Psychology, 45(3), 485.
Kossek, E. E., Pichler, S., Bodner, T., & Hammer, L. B. (2011). Workplace social support and
work-family conflict: A meta-analysis clarifying the influence of general and work-familyspecific supervisor and organizational support. Personnel Psychology, 64(2), 289.
doi:10.1111/j.1744-6570.2011.01211.x

177

Kossek, E. E., Ruderman, M. N., Braddy, P. W., & Hannum, K. M. (2012). Work–nonwork
boundary management profiles: A person-centered approach. Journal of Vocational
Behavior, 81(1), 112–128. doi:10.1016/j.jvb.2012.04.003
Kreiner, G. E. (2006). Consequences of work-home segmentation or integration: a personenvironment fit perspective. Journal of Organizational Behavior, 27, 485–507.
Kreiner, G. E., Hollensbe, E., & Sheep, M. (2009). Balancing borders and bridges: Negotiating
the work-home interface via boundary work tactics. Academy of Management Journal,
52(4), 704–730.
Kreyenfeld, M., & Hank, K. (2000). Does the availability of child care influence the employment
of mothers? Findings from Western Germany. Population Research and Policy Review,
19(4), 317–337. doi:10.1023/A:1026556309080
Lance, C. E., Dawson, B., Birkelbach, D., & Hoffman, B. J. (2010). Method effects,
measurement error, and substantive conclusions. Organizational Research Methods, 13(3),
435–455. doi:10.1177/1094428109352528
Lautsch, B. A., Kossek, E. E., & Eaton, S. C. (2009). Supervisory approaches and paradoxes in
managing telecommuting implementation. Human Relations, 62(6), 795–827.
doi:10.1177/0018726709104543
Lee, B., & DeVoe, S. (2012). Flextime and profitability. Industrial Relations, 51(2), 298–316.
Lee, M. D., Kossek, E. E., Hall, D. T., & Litrico, J.-B. (2011). Entangled strands: A process
perspective on the evolution of careers in the context of personal, family, work, and
community life. Human Relations, 1–23. doi:10.1177/0018726711424622
Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the three
dimensions of job burnout. Journal of Applied Psychology, 81(2), 123–133.
Lockwood, C. M., & MacKinnon, D. P. (1998). Bootstrapping the standard error of the mediated
effect. In Proceedings of the 23rd annual meeting of SAS Users Group International (pp.
997–1002). Cary, NC: SAS Institute.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A
comparison of methods to test mediation and other intervening variable effects.
Psychological Methods, 7(1), 83–104. doi:10.1037//1082-989X.7.1.83
MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the indirect
effect: Distribution of the product and resampling methods. Multivariate Behavioral
Research, 39, 99–128.

178

Major, V. S., Klein, K. J., & Ehrhart, M. G. (2002). Work time, work interference with family,
and psychological distress. Journal of Applied Psychology, 87(3), 427–436.
doi:10.1037//0021-9010.87.3.427
Manger, T., Wicklund, R., & Eikeland, O. (2003). Speed communication and solving social
problems. Communications, 28(2003), 323–337.
Maslach, C., Jackson, S. E., & Leiter, M. P. (1997). Maslach Burnout Inventory (Third Edition).
In C. P. Zalaquett & R. J. Wood (Eds.), Evaluating Stress: A Book of Resources (pp. 191–
218). The Scarecrow Press, Inc.
Matusik, S. F., & Mickel, A. E. (2011). Embracing or embattled by converged mobile devices?
Users’ experiences with a contemporary connectivity technology. Human Relations, 64(8),
1001–1030. doi:10.1177/0018726711405552
Mayer, R., & Schoorman, F. (1992). Predicting participation and production outcomes through a
two-dimensional model of organizational commitment. Academy of Management Journal,
35(3), 671–684.
Milliken, F. (1990). Perceiving and Interpreting Environmental Change: an Examination of
College Administrators’ Interpretation of Changing Demographics. Academy of
Management Journal, 33(1), 42–63. doi:10.2307/256351
Milliken, F., Martins, L., & Morgan, H. (1998). Explaining organizational responsiveness to
work-family issues: The role of human resource executives as issue interpreters. Academy
of Management Journal, 41(5), 580–592.
Morgeson, F. P., & Humphrey, S. E. (2006). The Work Design Questionnaire (WDQ):
developing and validating a comprehensive measure for assessing job design and the nature
of work. Journal of Applied Psychology, 91(6), 1321–39. doi:10.1037/0021-9010.91.6.1321
Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is mediated and mediation
is moderated. Journal of Personality and Social Psychology, 89(6), 852–63.
doi:10.1037/0022-3514.89.6.852
Nansen, B., Arnold, M., Gibbs, M., & Davis, H. (2010). Time, space and technology in the
working home: An unsettled nexus. New Technology, Work and Employment, 25(2), 136–
154.
Netemeyer, R., Boles, J., & McMurrian, R. (1996). Development and validation of work–family
conflict and family–work conflict scales. Journal of Applied Psychology, 81(4), 400–410.
Nippert-Eng, C. E. (1991). Constructing the home/work boundary: Mental categories in
everyday life. In Society for the Study of Symbolic Interaction. Cincinnati, OH.
Nippert-Eng, C. E. (1996). Home and Work. Chicago, IL: University of Chicago Press.
179

Nishii, L., Lepak, D., & Schneider, B. (2008). Employee attributions of the “why” of HR
practices: Their effects on employee attitudes and behaviors, and customer satisfaction.
Personnel Psychology, 61, 503–545.
O’Connor, M. A. (2005). Corporate social responsibility for work/family balance. St. John’s Law
Review, 79(4), 1–28.
O’Driscoll, M. P., Brough, P., Timms, C., & Sawang, S. (2010). Engagement with information
and communication technology and psychological well-being. In P. L. Perrewé & D. C.
Ganster (Eds.), New developments in theoretical and conceptual approaches to job stress
(research in occupational stress and well-being), Volume 8 (pp. 269–316). Emerald Group
Publishing Limited.
O’Hara, R. B., & Kotze, D. J. (2010). Do not log-transform count data. Methods in Ecology and
Evolution, 1(2), 118–122. doi:10.1111/j.2041-210X.2010.00021.x
Odle-Dusseau, H. N., Britt, T. W., & Greene-Shortridge, T. M. (2012). Organizational workfamily resources as predictors of job performance and attitudes: the process of work-family
conflict and enrichment. Journal of Occupational Health Psychology, 17(1), 28–40.
doi:10.1037/a0026428
OECD. (1996). The Knowledge-Based Economy. Paris.
OECD. (2004). OECD Country Note: Early Childhood Education and Care Policy in the
Federal Republic of Germany. Paris.
Offstein, E. H., Gnyawali, D. R., & Cobb, A. T. (2005). A strategic human resource perspective
of firm competitive behavior. Human Resource Management Review, 15(4), 305–318.
doi:10.1016/j.hrmr.2005.11.007
Oliver, C. (1991). Strategic responses to institutional processes. Academy of Management
Review, 16(1), 145–179.
Ollier-Malaterre, A., Rothbard, N. P., & Berg, J. M. (2013). When Worlds Collide in
Cyberspace: How Boundary Work in Online Social Networks Impacts Professional
Relationships. Academy of Management Review, 38(4), 645–669.
doi:10.5465/amr.2011.0235
Ollier-Malaterre, A., Valcour, M., Den Dulk, L., & Kossek, E. E. (2013). Theorizing national
context to develop comparative work–life research: A review and research agenda.
European Management Journal, 31(5), 433–447. doi:10.1016/j.emj.2013.05.002
Olson, D. H., McCubbin II, H., Barnes, H., Larsen, A., Muxen, M., & Wilson, M. (1983).
Families: What Makes Them Work (p. 320). Beverly Hills, CA: SAGE Publications.

180

Orthner, D. K. (1980). Families in Blue: A Study of Married and Single Parent Families in the
Air Force. Washington, DC.
Orthner, D. K., & Axelson, L. (1980). The effects of wife employment on marital sociability.
Journal of Comparative Family Studies, 11(4), 531–545.
Orthner, D. K., & Pittman, J. F. (1986). Family contributions to work commitment. Journal of
Marriage and Family, 48(3), 573–581. doi:10.2307/352043
Osterman, P. (1995). Work/family programs and the employment relationship. Administrative
Science Quarterly, 40, 681–700.
Pauls, K. (2013). German childcare allowance raises questions about working moms. CBC News.
Retrieved from http://www.cbc.ca/news/world/german-childcare-allowance-raisesquestions-about-working-moms-1.1373144
Payne, S. C., Cook, A. L., & Diaz, I. (2012). Understanding childcare satisfaction and its effect
on workplace outcomes: The convenience factor and the mediating role of work-family
conflict. Journal of Occupational and Organizational Psychology, 85(2), 225–244.
doi:10.1111/j.2044-8325.2011.02026.x
Perry-smith, J. E., & Blum, T. C. (2000). Work-Family Human Resource Bundles and Perceived
Organizational Performance. Academy of Management Journal, 43(6), 1107–1117.
Peus, C. (2006). Work-family balance? The case of Germany (No. 0025).
Piszczek, M. M., & Berg, P. (n.d.). Expanding the boundaries of boundary theory: Regulative
institutions and work-family role management. Human Relations.
doi:10.1177/0018726714524241
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common method
biases in behavioral research: a critical review of the literature and recommended remedies.
The Journal of Applied Psychology, 88(5), 879–903. doi:10.1037/0021-9010.88.5.879
Poelmans, S., & Beham, B. (2008). The moment of truth: Conceptualizing managerial work-life
policy allowance decisions. Journal of Occupational and Organizational Psychology, 81(3),
393–410. doi:10.1348/096317908X314865
Poelmans, S., & Sahibzada, K. (2004). A multi-level model for studying the context and impact
of work–family policies and culture in organizations. Human Resource Management
Review, 14(4), 409–431. doi:10.1016/j.hrmr.2004.10.003
Powell, G. N., Francesco, A. M., & Ling, Y. A. N. (2009). Toward culture-sensitive theories of
the work – family interface y, 616(December 2007), 597–616. doi:10.1002/job

181

Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects
in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36(4),
717–31.
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing Moderated Mediation
Hypotheses: Theory, Methods, and Prescriptions. Multivariate Behavioral Research, 42(1),
185–227.
Raley, S. B., Mattingly, M. J., & Bianchi, S. M. (2006). How Dual Are Dual-Income Couples?
Documenting Change From 1970 to 2001. Journal of Marriage and Family, 68(February),
11–28.
Ralston, D., & Flanagan, M. (1985). The Effect of Flextime on Absenteeism and Turnover for
Male and Female Employees. Journal of Vocational Behavior, 26, 206–217.
Rau, B. L., & Hyland, M. A. M. (2002). Role conflict and flexible work arrangements: The
effects on applicant attraction. Personnel Psychology, 55, 111–136.
Renaud, K., Ramsay, J., & Hair, M. (2006). “You’ve got e-mail!”... shall I deal with it now?
Electronic mail from the recipient’s perspective. International Journal of Human-Computer
Interaction, 21(3), 313–332.
Richardson, K. M., & Thompson, C. a. (2012). High Tech Tethers and Work-family Conflict: A
Conservation of Resources Approach. Engineering Management Research, 1(1), 29–43.
doi:10.5539/emr.v1n1p29
Richman, A. L., Civian, J. T., Shannon, L. L., Hill, E. J., & Brennan, R. T. (2008). The
relationship of perceived flexibility, supportive work–life policies, and use of formal
flexible arrangements and occasional flexibility to employee engagement and expected
retention. Community, Work & Family, 11(2), 183–197. doi:10.1080/13668800802050350
Robinson, J., & Godbey, G. (1997). Time for Life: The Surprising Way Americans Use their
Time. University Park: Penn State University Press.
Rothbard, N. P., Phillips, K. W., & Dumas, T. L. (2005). Managing Multiple Roles: WorkFamily Policies and Individuals’ Desires for Segmentation. Organization Science, 16(3),
243–258. doi:10.1287/orsc.1050.0124
Rothbard, N. P., & Ramarajan, L. (2009). Checking Your Identities at the Door? Positive
Relationships Between Nonwork and Work Identities. In L. M. Roberts & J. E. Dutton
(Eds.), Exploring Positive Identities and Organizations: Building a Theoretical and
Research Foundation. New York: Psychology Press.
Rousseau, D. M. (1990). Assessing organizational culture: The case for multiple methods. In B.
Schneider (Ed.), Organizational Climate and Culture (pp. 153–192). San Francisco, CA:
Jossey-Bass.
182

Ryan, A., & Kossek, E. E. (2008). Work‐life policy implementation: Breaking down or creating
barriers to inclusiveness? Human Resource Management, 47(2), 295–310. doi:10.1002/hrm
Schaufeli, W., & Bakker, A. (2004). Job demands, job resources, and their relationship with
burnout and engagement: A multi-sample study. Journal of Organizational Behavior, 25,
293–315.
Schor, J. B. (1993). The Overworked American: The Unexpected Decline of Leisure. New York:
Basic Books.
Schultze, U., & Boland Jr., R. J. (2000). Knowledge management technology and the
reproduction of knowledge work practices. Journal of Strategic Information Systems, 9,
193–212.
Seyler, D., Monroe, P., & Garand, J. (1995). Balancing work and family: The role of employersupported child care benefits. Journal of Family Issues, 16(2), 170–193.
Shockley, K. M., & Allen, T. D. (2007). When flexibility helps: Another look at the availability
of flexible work arrangements and work–family conflict. Journal of Vocational Behavior,
71(3), 479–493. doi:10.1016/j.jvb.2007.08.006
SHRM. (2010). 2010 Employee Benefits. Alexandria, VA.
SHRM. (2012). Technology and Its Impact on Employees During Nonworking Hours.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: New
procedures and recommendations. Psychological Methods, 7, 422–445.
Smith, A. (2011). 35% of American adults own a smartphone. Washington, DC.
Smith, A. (2013). Smartphone Ownership – 2013 Update. Washington, DC.
Somers, M. (1995). Organizational commitment, turnover and absenteeism: An examination of
direct and interaction effects. Journal of Organizational Behavior, 16(1), 49–58.
Spector, P. (1987). Method variance as an artifact in self-reported affect and perceptions at work:
myth or significant problem? Journal of Applied Psychology, 72(3), 438–443.
Stavrou, E., & Kilaniotis, C. (2009). Flexible Work and Turnover: an Empirical Investigation
across Cultures. British Journal of Management, 21, 541–554. doi:10.1111/j.14678551.2009.00659.x
Steel, R., & Ovalle, N. (1984). A review and meta-analysis of research on the relationship
between behavioral intentions and employee turnover. Journal of Applied Psychology,
69(4), 673–686.

183

Stolz, M. L. (2010). Work-Life in Germany. Boston, MA.
Stryker, S. (1980). Symbolic Interactionism: A Structrual Version. Menlo Park, CA:
Benjamin/Cummings.
Taylor, P. (2011, October). BlackBerry email glitch continues and spreads. Financial Times.
New York, NY.
The Local. (2010, June 28). Demand for day care on the rise. The Local.
Thomas, G., King, C., Baroni, B., Cook, L., Keitelman, M., Miller, S., & Wardle, A. (2006).
Reconceptualizing e-mail overload. Journal of Business and Technical Communication,
20(3), 252–287.
Towers, I., Duxbury, L., & Thomas, J. (2005). Time Thieves and Space Invaders: Technology,
Work and the Organisation Introduction: There is a time and a place for everything. In
Annual Critical Management Studies Conference (pp. 1–25). Cambridge.
Treas, J., & Drobnic, S. (Eds.). (2010). Dividing the Domestic: Men, Women, & Household Work
in Cross-National Perspective. Stanford, CA: Standford University Press.
Turel, O., Serenko, A., & Bontis, N. (2011). Family and work-related consequences of addiction
to organizational pervasive technologies. Information & Management, 48, 88–95.
doi:10.1016/j.im.2011.01.004
Turkle, S. (2007). Always-on/always-on-you: The tethered self. In J. Katz (Ed.), Handbook of
Mobile Communications and Social Change. Cambridge, MA: MIT Press.
Valcour, M., Ollier-Malaterre, A., Matz-Costa, C., Pitt-Catsouphes, M., & Brown, M. (2011).
Influences on employee perceptions of organizational work–life support: Signals and
resources. Journal of Vocational Behavior, 79(2), 588–595. doi:10.1016/j.jvb.2011.02.002
Van der Lippe, T. (2010). Women’s Employment and Housework. In J. Treas & S. Drobnic
(Eds.), Dividing the domestic: Men, women, & household work in cross-national
perspectives (p. 41). Stanford, CA: Standford University Press.
Van Iddekinge, C. H., Ferris, G. R., Perrewé, P. L., Perryman, A. A., Blass, F. R., & Heetderks,
T. D. (2009). Effects of selection and training on unit-level performance over time: a latent
growth modeling approach. Journal of Applied Psychology, 94(4), 829–843.
doi:10.1037/a0014453
Voydanoff, P. (1984). Work and Family: Changing Roles of Men and Women. Palo Alto, CA:
Mayfield.
Wayne, J. H., Casper, W. J., Matthews, R. a, & Allen, T. D. (2013). Family-supportive
organization perceptions and organizational commitment: the mediating role of work-family
184

conflict and enrichment and partner attitudes. The Journal of Applied Psychology, 98(4),
606–22. doi:10.1037/a0032491
Whittaker, S., & Sidner, C. (1996). Email overload: exploring personal information management
of email. In Proceedings of the SIGCHI Conference (pp. 276–283). Vancouver, BC.
Winkel, D. E., & Clayton, R. W. (2010). Transitioning between work and family roles as a
function of boundary flexibility and role salience. Journal of Vocational Behavior, 76(2),
336–343. doi:10.1016/j.jvb.2009.10.011
Wright, P., Gardner, T., Moynihan, L. M., & Allen, M. R. (2005). The relationship between HR
practices and firm performance: Examining causal order. Personnel Psychology, 58, 409–
446.
Wrohlich, K. (2005). The Excess Demand for Subsidized Child Care in Germany (No. 1515).
Zerubavel, E. (1993). The Fine Line: Making Distinctions in Everyday Life. Chicago: University
of Chicago Press.
Zerubavel, E. (1996). Lumping and splitting: Notes on social classification. Sociological Forum,
11(3), 421–433.

185