INDUSTRIAL RELATIONS, Vol. 42, No. 2 (April 2003).  2003 Regents of the University of California
Published by Blackwell Publishing, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington
Road, Oxford, OX4 2DQ, UK.
221
Blackwell Publishing LtdOxford, UKIRELIndustrial Relations0019-8676 2003 Regents of the University of CaliforniaApril 20034221000Original ArticleFamily-Friendly Management in Great BritainStephen J. Wood et al.
Family-Friendly Management in Great Britain:
Testing Various Perspectives
STEPHEN J. WOOD, LILIAN M. de MENEZES,
and ANA LASAOSA*
Five main perspectives on family-friendly management are differentiated by
their conceptions about the nature of such management and not just by its
assumed predictors. Multivariate analysis of the relationships among a set of
family-oriented practices shows that some but not all are used in a systematic
way. Regression analysis reveals that employers' adoption of family-friendly
approaches is explained by factors that span all five perspectives, but overall,
the organizational adaptation perspective fares best.
FAMILY-FRIENDLY MANAGEMENT INVOLVES EMPLOYERS having an
underlying commitment to help employees obtain a balance between work
and family obligations. This commitment or orientation is expressed in
management policies, so there should be some pattern to the provision of a
range of family-friendly practices. For example, we expect practices concerned
with childbirth to coexist with those related to child rearing.
The recent wave of research in the United States on family-friendly
practices concentrated on the predictors of their adoption rather than on
the relationships among them (Goodstein 1994, 1995; Milliken, Dutton,
and Beyer 1990; Ingram and Simons 1995; Osterman 1995). Researchers
sought to test a particular theory or adjudicate between theories based
on characterizing them according to their determinants of family-friendly
* The author's affiliations are, respectively, Institute of Work Psychology, University of Sheffield; Cass
Business School, City university, London; and Centre for Economic Performance, London School of
Economics. E-mail: s.j.wood@sheffield.ac.uk, mas01ld@gold.ac.uk, and alasaosa@yahoo.com. The United
Kingdom's Economic and Social Research Council funded this research (Grant No. R000238112). We
would like to thank Catherine Cassell, Alan Felstead, Suzan Lewis, three anonymous referees, and the
editors for their comments on an earlier draft of this article. The empirical research is based on data
from the 1998 Workplace Employee Relations Survey (WERS98). This survey is jointly sponsored by the
Department of Trade and Industry of the United Kingdom, the Advisory, Conciliation and Arbitration
Service, the Economic and Social Research Council, and the Policy Studies Institute. The National
Centre for Social Research was commissioned to conduct the survey field work on behalf of the sponsors.
WERS98 is deposited and available from the Data Archive at the University of Essex, UK. Neither the
sponsors nor the Data Archive have any responsibility for the analysis or interpretation of the material
contained in this article.
222 / STEPHEN J. WOOD ET AL.
practices. Typically, the number of family-friendly practices used was
regressed on the determinants associated with each theory. In so doing,
researchers attributed equal status to family-friendly practices and implied
that a managerial orientation underlay their use.
Different theories of family-friendly management, however, can be differentiated
by what is being predicted and not simply by their predictors. To
test these theories requires an analysis of both the nature of employers'
involvement in work-family management (the dependent variable) and its
determinants (the independent variables).
In this article we first identify five perspectives on family-friendly management:
(1) institutional theory, (2) the organizational adaptation variant
of institutional theory, (3) the high-commitment perspective, (4) the equalopportunity
perspective, and (5) the situational perspective. We show that
in their strong versions they yield different hypotheses and have conflicting
concepts of family-friendly management. In the institutional theory, familyfriendly
management, as defined above, is assumed to be an identifiable and
integrated phenomenon, and hence there is a coherency to the use of familyfriendly
practices that reflects an underlying managerial orientation. In
the high-commitment and equal-opportunity perspectives, such a familyfriendly
management is part of a bigger approach. By contrast, from the
situational perspective, the provision of family-friendly practices at best
would be fragmented, whereas under the organizational adaptation perspective,
the expected pattern of associations is an open question.
Second, we assess which of the five perspectives, either alone or in combination
with others, best predicts the nature of any adoption of familyfriendly
practices and which are the determinants of this adoption. We use
data from Great Britain's Workplace Employee Relations Survey of 1998
(WERS98). The study is in two stages:
1. Multivariate analyses of the association among a set of familyorientated
practices in order to establish whether family-friendly
management is an integrated concept that is summarized in a
single dimension
2. Given coherence in the provision of family-friendly practices, an
assessment of the relative importance of the major predictors of
the various theories
The first stage has not been undertaken by U.S. researchers--although
Wood (1999a) conducted similar analysis of Osterman's U.S. data. This
article thus brings some British data to bear on a problem that until now
largely has been addressed with U.S. data.
Family-Friendly Management in Great Britain / 223
Five Perspectives on Family-Friendly Management
Demographic changes such as the increased participation of women in
the workplace have highlighted the importance of achieving a work-family
balance. This has created institutional pressure on employers to "assume a
more active role in helping employees balance work and family demands"
(Goodstein 1994:356). Employers have begun to respond by offering workfamily
practices or benefits. This chain of reasoning is captured most
directly in the institutional perspective.
Institutional theory. According to institutional theory, the adoption of
family-friendly practices is a way of conforming to normative pressures
from the wider society and of institutionalizing family-friendly management.
Consequently, underlying the adoption of all practices ought to be an
orientation on the part of management toward resolving work-family conflicts.
Variability in the adoption of such practices is explained by the extent
to which the maintenance of social legitimacy is salient to the organization.
Large private firms and public-sector organizations, being more visible, are
assumed to be more compelled to conform to societal pressures; also, they
may have more to gain by doing so (Oliver 1991; Goodstein 1994). It has
been argued that the pressures to respond to the specific work-family issue
might be greater in some industries, e.g., health care (Morgan and Milliken
1992:228). Here we thus test the following hypotheses of institutional theory:
Hypothesis 1. Family-friendly practices coexist and the pattern of associations
among them is reducible to a unidimensional structure.
By unidimensional structure we mean that the correlation between practices
reflects a common concept; that is, they reflect in this case a single form of
management, a holistic family-friendly management.
Hypothesis 2. Large organizations, public-sector organizations, and health
care organizations are more likely to conform to institutional pressures to
adopt family-friendly management.
Organizational adaptation theory. Organizational adaptation theory is
best seen as an extension of institutional theory in order to allow for the
strategic choice exercised by management (Goodstein 1994, 1995). Its starting
point is institutional theory--the premise that organizations must
respond to demands embodied in regulations, norms, laws, and social
expectations--but it departs from institutional theory in two ways. First, it
assumes that organizations do not passively conform to normative pressures
and that management has discretion over its response to environmental
224 / STEPHEN J. WOOD ET AL.
challenges (Oliver 1991; Daft and Weick 1984). Management may even choose
to ignore or actively oppose societal pressures to balance work and family.
Second, organizations are assumed to differ in how they scan and define
these apparent external pressures. Thus weight is given to how management
perceives and interprets or "recognizes" the institutional environment.
The normative pressures arising from the changing structure of the family
inevitably impose constraints on employers' actions, but under this perspective,
managers can interpret these issues in different ways. Some may
treat family-work pressures as given and even assimilate them into a new
high-commitment management. Others may interpret the issues as pertaining
to women or particularly women with young children. Some may ignore
the normative pressures by regarding family-related issues as the employee's
private responsibility and thus of little direct concern to the organization.
Accordingly, family-friendly management is, under this perspective, not
necessarily a one-dimensional concept because it suggests that clusters or
subsets of practices may well exist. For example, there may be no association
between the two types of family-oriented practices that are commonly
differentiated--those which provide a substitute caregiver for the employee
and those which create flexibility in the timing and location of work so that
the employee can more readily accommodate family demands (Bailyn
1993:67 and Bond, Galinsky, and Swanberg 1998).
Goodstein (1994:363­5) bases his index of "responsiveness to institutional
pressures for involvement in work-family issues" on this distinction
between child care and flexible working practices. However, he neither
assumes nor tests a general relationship between these practices. He in fact
assumes a connection only at high levels of responsiveness, since his index
grades employers' responses according to Oliver's general fourfold classification
of responses--acquiescence, compromise, avoidance, and defiance--
and acquiescence is measured by whether the organization adopts at least
one or more of both types of practices. Compromise is when the organization
provides only one type but adopts multiple kinds of it. Avoidance is
when the organization provides one practice only, which can be of either
type. Finally, defiance is when the organization provides none of the practices.
Goodstein thus prespecifies the link between practices and the underlying
orientation of management toward assumed societal pressures.
Nevertheless, taken to its logical extreme, the organizational adaptation
perspective implies that the nature of any family-friendly management is an
open empirical question.
This perspective also broadens the range of predictors of the provision of
family-friendly practices. Technical factors, competitive pressures, and local
situational factors enter management's deliberations because the interpretation
Family-Friendly Management in Great Britain / 225
process is not simply a matter of ethical considerations. Consequently, in
addition to the predictors of the institutional theory, in the organizational
adaptation perspective we have (1) the values of senior management, (2) the
information senior management has on institutional pressures (particularly
through its knowledge of its own employees, past, potential, and present),
and (3) the perceptions of the economic outcomes associated with the
provision of family-friendly practices. It is therefore thought that employers
are more likely to be family-friendly if they value their employees having
a healthy work-family balance, consult their employees regularly about
their needs and concerns, and associate family-oriented management with
tangible performance effects (see especially Goodstein 1995:1660­1;
Milliken, Dutton, and Beyer 1990:101; Morgan and Milliken, 1992:230­1).
In this study we will test the core of the organizational adaptation perspective
by
Hypothesis 3. The more highly management values employees having a balance
between work and the family, the greater will be its adoption of familyfriendly
management.
Hypothesis 4. The more management is aware of its workforce's needs, the
greater will be its adoption of family-friendly management.
The organizational adaptation literature also accepts that organizations'
responses reflect variations in the intensity of pressure at the local level. For
example, managements faced with problems of absenteeism arising from
family problems are more likely to be concerned about family-friendly
issues. Pressures employees bring to bear on senior management within an
organization also may affect the prominence given to societal pressures.
Hence, if women are the constituency that makes the strongest demands for
work-family accommodation, organizations with a high proportion of
females in the workforce will be more likely to provide work-family facilities.
Also, employees who are in a strong position because of their transferable
skills (e.g., professional staff) may articulate their views and influence
management to a greater extent than less powerful employees (Goodstein
1995:1661). Thus, organizations with a high proportion of professional staff
are more likely to practice family-friendly management. We thus test
Hypothesis 5. The higher the proportion of the organization's employees who
are women, the greater will be its adoption of family-friendly management.
Hypothesis 6. The higher the skill level, the proportion of professional employees,
and the proportion of managerial employees in the organization, the
greater will be its adoption of family-friendly management.
226 / STEPHEN J. WOOD ET AL.
In its strongest form, the organizational adaptation approach assumes
that managements have near-complete freedom in interpreting issues and
scanning the environment. We would expect the organization's reaction to
external pressures to depend on its managers' concern for their employees'
work and family balance. This therefore would moderate the relationship
between organizational size and family-friendly management. Accordingly,
we will test
Hypothesis 7. The effect of size in family-friendly management is intensified
when management values a balance between work and the family.
The organizational adaptation approach could be taken further beyond
institutional theory by incorporating the strategy of the organization. Using
Porter's (1985) distinction between cost-minimization and high-quality
strategies, Milliken, Dutton, and Beyer (1990:98) argue that organizations
adopting the high-quality strategy are more likely to see work-family matters
as a strategic concern. The question then is whether the organization's
strategy is just an additional factor affecting the priority that the dominant
coalition gives to family-work issues, or whether it is the primary determinant
of family-friendly management (as it would be in a pure strategic-choice
model). This entails testing
Hypothesis 8. Organizations adopting high-quality strategies are more likely
to adopt family-friendly management.
The high-commitment perspective. The high-commitment perspective1
as developed by Osterman (1995) hypothesizes that the rise of highcommitment
management explains the increasing adoption of work-family
programs. However, a core element of the human resources management
debate (see, for example, Appelbaum and Batt 1994; Huselid 1995;
MacDuffie 1995; Wood 1996, 1999b; Wood and Albanese 1995) is whether
this high-commitment management is universally applicable. If its adoption
is contingent on the business strategy, then high-commitment management
will mediate the relationship between it and family-friendly management.
This effectively amounts to an extension of Milliken, Dutton, and Beyer's
Porterian hypothesis.
Osterman does not comment on the possible contingent nature of highcommitment
management in his family-friendly article, but he associates
1
There are several terms currently being used to describe a form of management similar to highcommitment
management (see Walton 1985), including high-involvement management (Lawler 1986)
and high-performance management (e.g., Appelbaum and Batt 1994:5). The term high commitment will
be used here because it corresponds to Osterman's terminology.
Family-Friendly Management in Great Britain / 227
high-commitment management with "performance gains" (Osterman
1995:698) in a way that suggests that it is increasingly generally relevant (see
also Kochan and Osterman 1995; Capelli et al. 1997). This implies that
high-commitment management is a main predictor. Thus
Hypothesis 9. Managements that place a high value on obtaining employee
commitment and are implementing high-commitment management are more
likely to adopt family-friendly management.
Taking the argument to its extreme, high-commitment and familyfriendly
practices would reflect a single underlying managerial approach.
We will thus test
Hypothesis 10. Family-friendly practices and high-commitment practices coexist,
and the pattern of associations among them reduces to a one-dimensional
structure.
The equal-opportunity perspective. The equal-opportunity perspective
refers to the use of family-friendly practices to remove discrimination particularly
on the basis of gender, race, or age. Reconciliation of employment
and family life is central to any discussion of equal opportunities for men
and women. Several authors have claimed that meaningful family-friendly
management involves "confronting the issues of discrimination--based not
just on gender, but also on race, ethnicity, and other kinds of differences"
(Hall 1990:69). Unless this is done, it is claimed, organizations will treat
family-friendly practices as part of their fringe benefits and are only likely
to introduce them in a half-hearted and piecemeal way, not as part of
broader program that aims to reduce unequal opportunities (Gonyea and
Googins 1996:69; Lewis 1997:15). The minimal conclusion of this argument is
Hypothesis 11. Managements that have equal-opportunity policies are more
likely to adopt family-friendly management.
Taking the argument further, it may be that equal-opportunity and
family-friendly practices will reflect a single managerial approach. Thus
Hypothesis 12. Family-friendly practices and equal-opportunity practices
coexist in a unidimensional structure.
Finally, as part of a switch in emphasis from removing disadvantages to
valuing differences, family-friendly practices and equal-opportunities policies
have been linked with high-commitment management. Bailyn (1993:87)
argued that equality and diversity agendas imply that any effective highcommitment
management must be extended from employee involvement
to embrace issues of working time. It should be based on the belief that
228 / STEPHEN J. WOOD ET AL.
"personal time is a legitimate employee need." This argument is normative,
but it does imply that any serious move toward any one of the triad--
equal-opportunity management, high-commitment management, and
family-friendly management--will involve the others. All will reflect the
same fundamental change in employers' orientations. Accordingly, we will
test the strongest version of this perspective:
Hypothesis 13. Family-friendly practices, equal-opportunity practices, and
high-commitment practices coexist in a one-dimensional structure.
The situational perspective. The situational perspective, or what Osterman
(1995) calls the "practical response perspective," is a complete break from
institutional theory. Under this perspective, managements adapt to local
circumstances rather than to societal normative pressures. On the personnel
front, managements adapt to the idiosyncrasies of their workforce and
other local labor market pressures. Accordingly, we would expect the characteristics
of the workforce to be the major predictors of the degree of
family-friendly management. More specifically, we would assume gender
composition to be important. Thus hypothesis 5--the higher the proportion
of the organization's employees who are women, the greater is the adoption
of family-friendly management--becomes the main hypothesis of this
perspective.
In addition, employers who perceive that labor problems such as absenteeism
and stress are related to family problems will respond by adopting
family-friendly practices. Since one expects such problems to be related to
the age of the employees' children, we can formulate the following hypothesis:
Hypothesis 14. The greater the proportion of employees with young children,
the greater will be the adoption of family-friendly practices.
A tight labor market also may prompt employers to offer family-friendly
practices as a way of attracting potential candidates. Thus
Hypothesis 15. The greater the difficulties an organization experiences in
filling vacancies, the greater will be the adoption of family-friendly practices.
Under the situational perspective, we would not necessarily expect
employers to think of the family "problem" as a holistic issue. Managers
would be more likely to adopt family-friendly practices in a piecemeal way.
Consequently, a test of hypothesis 1--that the association between familyfriendly
practices is reducible to a unidimensional structure--is also a test
of this perspective. A rejection of this hypothesis would imply an ad hoc or
fragmented adoption of family-friendly practices.
Family-Friendly Management in Great Britain / 229
A coherent response across organizations is only likely if it is widely
accepted that family-friendly practices have positive payoffs. If there are any
economies of scale in the provision of family-friendly practices, then we
might view the size of the organization as another predictor.
We have shown that the five perspectives can be differentiated by their
conceptions of family-friendly management and not simply by their predictions.
In strong institutional theory, family-friendly management is considered
a one-dimensional concept, whereas in the situational perspective, this is
highly unlikely. In the high-commitment management and equal-opportunity
perspectives, family-friendly management is part of a broader managerial
approach. Finally, in the organizational adaptation perspective, the form of
any collective adoption of family-friendly practices is an open question.
In this study we first test the various assumptions about the nature of
family-friendly management. Subsequently, we test the predictors associated
with each perspective. If it is found that the provision of family-friendly
practices is ad hoc, the analysis must be limited to predicting the adoption
of individual practices. In the less extreme case, where there is still no
unified family-friendly management but only discrete elements of it, the
analysis will focus on each element separately. Finally, if family-friendly
management is part of a broader high-commitment approach and/or equalopportunity
policy, the task will be to predict this overall approach.
We aim to assess the various perspectives outlined above by analyzing the
following:
1. Whether family-friendly practices are provided in a systematic
way
2. Whether there is ad hoc provision of family-friendly practices
3. Whether there are different dimensions to family-friendly
management
4. The relationship between the provision of family-friendly, highcommitment
management and equal-opportunity practices
5. The relative importance of institutional factors, values, employers'
awareness of their employees, high-commitment management,
equal-opportunity policy, and local situational factors in explaining
the adoption of family-friendly practices
We address the first four questions using latent variable analysis. We then
use regression analysis to assess the relative importance of the different
factors. Table 1 summarizes the main hypotheses, their linkages to the five
perspectives, and the test procedures that we adopt.
230 / STEPHEN J. WOOD ET AL.
TABLE 1
RESEARCH HYPOTHESES BY THEORETICAL PERSPECTIVE
Hypotheses Perspective Testing Method
1. Family-friendly practices coexist and the associations
reduce to a unidimensional structure.
Institutional Latent variable
analysis
2. Large organizations, public-sector organizations,
and health care organizations are more likely
to conform to institutional pressures to adopt
family-friendly practices.
Institutional Regression
analysis
3. The greater the extent to which management values
employees having a balance between work, the greater
the use of family-friendly practices.
Organizational
adaptation
Regression
analysis
4. The more management is aware of its workforce's needs,
the greater the adoption of family-friendly practices.
Organizational
adaptation
Regression
analysis
5. The greater the proportion of women in the workforce,
the greater the use of family-friendly practices.
Organizational
adaptation/situational
Regression
analysis
6. The higher the skill level and the proportion of
professional employees in the organization, the greater
the adoption of family-friendly practices.
Organizational
adaptation
Regression
analysis
7. The effect of size on family-friendly management is
intensified when management values having a balance
between work and family life.
Organizational
adaptation
Regression
analysis
8. Organizations adopting high-quality strategies are more
likely to adopt family-friendly practices.
Organizational
adaptation
Regression
analysis
9. Management placing a high value on employee
commitment and implementing high-commitment
practices are more likely to adopt family-friendly
practices.
High commitment Regression
analysis
10. Family-friendly practices and high-commitment
practices coexist and the associations among them
are reducible to a unidimensional structure.
High commitment Latent
variable
analysis
11. Managements that are oriented toward
equal-opportunity practice are more likely to
adopt family-friendly practices.
Equal opportunity Regression
analysis
12. Family-friendly practices and equal-opportunity
practices coexist and the associations among them are
reducible to a unidimensional structure.
Equal opportunity Latent
variable
analysis
13. Family-friendly, equal-opportunity, and highcommitment
practices coexist and the associations
among them are reducible to a unidimensional
structure.
Equal opportunity Latent variable
analysis
14. The greater the proportion of employees with young
children, the greater the use of family-friendly practices.
Situational Regression
analysis
15. The greater the difficulties in filling vacancies that
an organization experiences, the greater the use of
family-friendly practices.
Situational Regression
analysis
Family-Friendly Management in Great Britain / 231
The Data Set
Workplace Employee Relations Survey of 1998 (WERS98). The WERS98
data set is the fourth in a series of surveys, the Workplace Industrial Relations
Series (WIRS), that is aimed at achieving a "better understanding of
the processes which underlie employment relationships" through a largescale
survey of a "broad range of industrial relations and employment
practices" (Millward et al. 1992). Its unit of analysis is the workplace or
establishment.
Three research instruments make up WERS98. These are a structured
interview with the senior manager at the site responsible for employee relations,
a structured interview with an employee representative (where such a
person existed), and a questionnaire completed by a random sample of up
to 25 employees in the majority of workplaces. The analysis in this article
uses data from both the management and employee interviews.
Field work was carried out between October 1997 and June 1998. Workplaces
employing 10 or more employees (excluding agriculture, fishing, and
deep coal mining) were randomly sampled from the official Inter-Departmental
Business Register (of 1997) that is used for most British government
surveys of employers. The sampling frame was stratified by workplace size
and industry sector. A sample of 2191 workplaces was achieved, representing
a response rate of 80.3 percent for the senior management interview.
Managers at 85.8 percent of the 2191 workplaces agreed to cooperate with
the employee survey. The employees who filled out the employee questionnaire
were sampled randomly from a list provided by management at these
workplaces. In practice, there were nil returns of employee questionnaires
from 4.4 percent of workplaces, so 81.4 percent of the workplaces in the
management survey (totaling 1783) participated in the employee survey. Of
the 44,283 questionnaires sent to individual employees, 28,215 were
returned, for a response rate of 64 percent.
In the analysis that follows we use the weights provided in the WERS98
data set whenever we investigate effects on the population or make any inference
regarding Great Britain. In doing so, we correct the sample bias that
results from the design of the sample, which led to an overrepresentation of
large workplaces and certain industries [see Airey et al. (1999:4­8) for more
details of the sample and (1999:88­92) for the weighting of the data].
Variables used in the study. The family-friendly practices in the data set
all relate to nonmanagerial employees. They include both flexible working
practices and child care assistance. More specifically, there is information
on whether employees are entitled to parental leave, work at home,
232 / STEPHEN J. WOOD ET AL.
term-time-only contracts, part-time work (once having been full time), job
sharing, use of a workplace nursery, and a financial subsidy toward child
care. There is no information about elder care arrangements or employee
assistance with retirement or drug or alcohol abuse.2
The high-commitment practices in the data set, which will be used in the
subsequent analysis, are quality circles, functional flexibility, teamwork,
suggestion schemes, induction, interpersonal skills training, team briefing,
information disclosure, and appraisal.
The measures of equal-opportunity practices are as follows (all phrased
as in the interview schedule): maintenance of workplace records on ethnic
origin of employees; collection of statistics on posts held by men and
women; monitoring of promotions by gender, ethnicity, and so on; revision
of selection and other procedures to identify indirect discrimination; revision
of relative pay of different groups; and adjustments to accommodate
disabled employees.
In WERS98, there are several measures associated with the institutional
perspective: the employment level of the workplace and the employment
level of the total organization of which the workplace is a part, whether the
workplace is in the public or private sector, and the industry group to which
it belongs.
A variable of particular relevance to the organizational adaptation perspective's
hypothesis 3 is the employers' concern for their employees' balance
between work and family life. Another set of variables is related to the
employer's ability to be aware of their employees' needs (hypothesis 4): the
degree to which management consults with the workforce, the existence of
a personnel department, and the availability of employee voice through a
recognized trade union.
Hypothesis 5 is tested using the proportion of females in the workforce.
Hypothesis 6 is tested using the extent of education of the workforce, the
proportion of employees who are technical and professional, and the proportion
of employees who are managerial. We lack a precise measure of
whether the workplace pursues a high-quality strategy, but we use a measure
of total quality management (TQM) as a proxy to test hypothesis 8.
Four measures related to the high-commitment perspective are included:
the extent to which management accords importance to employee commitment,
the importance given to human resources management in the planning
process, management's orientation toward high commitment (see below for
the development of the measure of this based on high-commitment practices),
2
An influence on the survey design was the Labour Party's encouragement of family-friendly practices
in its Manifesto of 1995 and the way that this was associated with child care.
Family-Friendly Management in Great Britain / 233
and the extent to which the employees in the largest occupational group
have enriched jobs. These will be used to test to hypothesis 9.
The key variable for the equal-opportunity perspective is the extent of
equal opportunity measured by a scale (as shown below) based on all the
equal-opportunity practices.
The measures of situational factors, apart from the proportion of females
in the workforce, are the proportion of employees who have children of
preschool age and the proportion of employees who have children below the
age of 19 (hypothesis 14). The extent to which the employer is faced with
labor market problems is measured by whether there have been difficulties
filling vacancies (in the workplace) in the year previous to the data collection
(hypothesis 15). We lack variables indicating the level of absenteeism due to
family problems or the perceived benefits from family-friendly practices.
Finally, we include the following controls in our analysis: whether the
workplace is part of a larger organization, employment change, the proportion
of part-timers, and some additional occupational composition measures
such as the percentage of clerical and secretarial employees.
Table 2 outlines the independent variables that are used in the regression
analysis.
The Analysis of Family-Friendly Management
Latent variable analysis is used here to make inferences from the relationships
among the practices, be these family-friendly, high-commitment, and
equal-opportunity practices. Latent variable models, of which factor analysis
is the oldest and best known, assess whether the association between a
set of manifest or observed variables (in this case management practices)
can be explained by a common (unobserved) factor.3
Different models can
be used, depending on the type of data, because our manifest variables--
family-friendly, equal-opportunity, and high-commitment practices--are
binary, latent trait, and latent class models. The difference between these
models is whether the underlying unobservable variable is continuous (in
the latent trait model) or categorical (in the latent class one). The logit-probit
(latent trait) model (Bartholomew and Knott 1999) is used to test hypotheses
3
See Bartholomew and Knott (1999:1­39) for an account of latent trait models, and Bartholomew,
de Menezes, and Tzamourani (1997) and the Appendix in Wood (1999c:411­4) for the logit-probit
model used here. For other applications of the latent trait model, see Bartholomew, de Menezes, and
Tzamourani (1997), de Menezes and Bartholomew (1996), Wood (1996, 1999a, 1999c), Wood and
Albanese (1995), and Wood and de Menezes (1998).
234 / STEPHEN J. WOOD ET AL.
TABLE 2
INDEPENDENT VARIABLES USED IN REGRESSION ANALYSIS OF FAMILY-FRIENDLY MANAGEMENT
Institutional perspective
Employment level
Total number of full- and part-time employees in the workplace.
Size of the total organization of which the workplace is a part
Categorical variable indicating the total number of employees in the company of which the
establishment is a part. (In the case of the single-site workplace, this is the same as the employment
level.)
Industry
Nine different industry groups, each with workplaces in the private and public sector.
Public
Whether the workplace belongs to the public or private sector, regardless of its industry group.
Organizational adaptation perspective
Employers' concern for their employees' balance between work and family life
The management's orientation toward family issues, measured on the basis of a question in which
respondents were asked, "It is up to individual employees to balance their work and family
responsibilities?" (1 = strongly agree, 2 = agree, 3 = neither agree nor disagree, 4 = disagree, 5 = strongly
disagree). The variable has been recoded binary one. It is positive for values 3 to 5 (16% of workplaces).
Consultative approach
An index of the extent to which management consults with its workforce, created from a question asking
whether the organization had employee-management committees or regular meetings of employees and
managers to discuss five issues: health and safety, productivity, training, technology and work
organization, work roles or job descriptions. A latent trait model was fitted to the data on the set of five
types of committees/meetings (% of G2
= 77.24). The variable included in the regressions is the resulting
latent trait score.
Personnel department
There is a personnel specialist who spends at least 25% of his or her time on personnel matters.
Union representation
At least one union is recognized at the workplace.
The percentage of female employees
Proportion of regular (full- and part-time) employees who are female.
Educated workforce
Positive if more than half the employees have a university degree.
Proportion of technical and professional employees
Proportion of employees in the workplace belonging to the professional and technical occupational
group.
Proportion of managerial employees
Proportion of employees in the workplace who are managerial.
Quality strategy (total quality management)
Latent trait score resulting from a model with the following practices: self-inspection, keeping records
on faults/complaints, customer surveys, nonconfidential quality records, and training in quality control.
(% G2
= 73.51)
Family-Friendly Management in Great Britain / 235
High-commitment perspective
Importance accorded to employee commitment
Based on the agreement index to the statement, "Employees are led to expect long-term employment in
this organization," and recoded so that 1 means strongly disagree, 2 disagree, 3 neither agree nor
disagree, 4 agree, and 5 strongly agree.
Importance given by management to human resources
Variable measuring if (1) the strategic plan includes employee's development and (2) the person
responsible for employee relations was involved in the plan design. It can take the values 0, 1 or 2.
High-commitment management
A classification of workplaces into three different categories (minimal, partial, and full highcommitment
management). This grouping is the result of a latent class analysis of nine practices found
to belong together: teamworking, functional flexibility, quality circles, suggestion schemes, team
briefings, induction procedure, training in human relations, and information disclosure (see Table 7).
Enriched jobs
Average of three items (job variety, job discretion, and job control) from the manager's questionnaire.
Each item is measured on a four-point scale, from none to a lot.
Equal-opportunity perspective
Equal-opportunity orientation
An index of the extent to which equal-opportunity policies are implemented, based on the latent trait
analysis of equal opportunity practices described in the text (model 8 in Table 5).
Situational factors perspective
Difficulties in filling vacancies
Management has encountered difficulties in filling vacancies over the previous year.
Proportion of employees who are parents with children aged 4 or less
Proportion of all employees who are parents with children aged 4 or less.
Proportion of employees who are parents with children aged 18 or less
Proportion of all employees who are parents with children aged 18 or less.
Control variables
Part of a larger organization
Measures whether the workplace is an independent or stand-alone site (1, if the organization is part of
a wider organization or enterprise).
Change in employment size
Measures the difference between the number of employees (full- and part-time) employed at the
workplace at the time of the interview and 1 year before.
Additional occupational composition measures
Variables indicating the proportion of employees in the workplace belonging to the following
occupational categories: routine and skilled, operational, clerical and secretarial.
Number of part-timers
Proportion of employees who work part time.
TABLE 2 CONTINUED
236 / STEPHEN J. WOOD ET AL.
1, 10, 12, and 13. This is consistent with the definition of unidimensionality
as "the existence of one latent trait underlying the data" (Hattie 1985:157).
The goodness of fit of the latent trait model is judged by two methods.
First, we look at overall fit measures. One of them is the percentage of G2
(the log-likelihood ratio statistic for complete independence) that compares
the fit with what it would be if there were an association. The closer this
statistic is to 100 percent, the better is the fit of the model. Another measure
is the chi-square for all observed response patterns. Second, we assess the fit
by comparing the observed (O) and expected (E) response patterns for pairs
and triplets of items. This is done by constructing the statistic (O - E)2
/E
for each response pattern. The closer this is to zero, the better is the fit of
the model. When a significant number of these statistics are large, there are
residual associations between items that are not attributable to the latent
variable(s). Furthermore, if the same observable items (e.g., family-friendly
practices as in our main analysis) are found in the problematic pairs and
triplets, it suggests that they are the cause of bad fit. Such a pattern would
contradict the basic assumption of the model: that our latent variable (familyfriendly
management) is the only source of correlation among observed
family-friendly practices.
Testing hypothesis 1: The latent variable analysis of family-friendly
practices. Table 3 (columns 2 and 3) records the relative adoption of particular
family-friendly practices, revealing that there is considerable variation
in adoption. No practice is adopted in the majority of the workplaces
across the whole economy. The most frequently available practice is the
entitlement to work part time (46 percent of all workplaces in the economy,
58 percent of the sample); this is followed by parental leave (34 and 43
percent) and job sharing (28 and 38 percent). The provision of a workplace
nursery and child care subsidies is very rare, the figures for the whole
economy being 3 and 4 percent, respectively (8 and 9 percent in the sample).
TABLE 3
PROVISION OF FAMILY-FRIENDLY PRACTICES FOR NONMANAGERIAL EMPLOYEES
In WERS98
(Unweighted Proportion) In the Economy
Parental leave 43 34
Working from home 18 13
Term-only contracts 20 16
Working part time 58 46
Job sharing 38 28
Workplace nursery 8 3
Child care subsidies 7 4
Family-Friendly Management in Great Britain / 237
As Table 4 shows, chi-square tests on all the pairs of family-friendly
practices imply coexistence within the data. The associations among variables
turned out to be positive in all cases, including that between workplace
nursery provision and child care subsidy. Contrary to what one might
expect, we find no substitution effect between the provision of these two
practices. That is, a workplace that has one of the two practices is more
likely to have the other.
The latent trait analysis of all family-friendly practices, which tested
whether the associations among family-friendly practices are reducible to a
unidimensional structure, was not successful. All goodness-of-fit measures
that we used rejected these hypotheses. Table 5 shows that (O - E)2
/E is
above 4 in eight pairs of items in this model (1). The two child care items--
provision of a workplace nursery and child care subsidy--are responsible
for these high residual associations.
We thus tested whether this reflects the existence of two underlying variables,
perhaps one relating to flexible working practices and the other to
child care facilities by fitting a two-factor model. The results displayed in
Table 5 (model 2) do not support this hypothesis: The fit does not improve,
and the same child care items are the problematic pairs and triplets. That is,
such items are associated independently of the variables being measured and
could themselves be conceived of as a single variable. However, combining
the two child care items into one does not solve the problem. When provision
of a workplace creche and child care subsidy are removed from the
model, the fit improves dramatically (Table 5, model 3), and all (O - E)2
/E
statistics are nearly zero. This suggests that the relationship between the
family-friendly practices associated with flexible working arrangements is
explained by a common factor, but the provision of facilities and/or financial
support for child care is discrete from this. We thus reject our first
TABLE 4
ASSOCIATION BETWEEN FAMILY-FRIENDLY PRACTICES: CHI-SQUARE TESTS
Parental
Leave
Working
from Home
Term-Only
Contracts
Part
Time
Job
Sharing
Workplace
Nursery
Parental leave
Working from home 62.09
Term-only contracts 70.04 86.7
Part time 207.6 117.4 170.4
Job sharing 227.9 157.5 201.1 561.7
Workplace nursery 28.7 63.6 142.5 42.2 109.4
Child care subsidies 36.9 60.9 7.3 34.6 89.1 92.6
NOTE: All the chi-square tests are significant at 0.001 level.
238/STEPHENJ.WOODETAL.
TABLE 5
DISCRIMINATION COEFFICIENTS FROM LATENT TRAIT MODELS OF FAMILY-FRIENDLY, HIGH-COMMITMENT, AND EQUAL-OPPORTUNITIES PRACTICES
Model Number 1 2 3 4 5 6 7 8
       
Number of factors 1 2 1 1 1 1 1 1
Family-friendly practices
Parental leave 0.74 0.61 0.41 0.73 0.77 0.70 0.72
Working from home 0.77 0.64 0.42 0.75 0.74 0.67 0.67
Term-only contracts 0.80 0.67 0.44 0.80 0.77 0.71 0.71
Working part time 0.94 0.78 0.52 0.95 0.90 0.78 0.79
Job sharing 0.96 0.80 0.53 0.95 0.91 0.84 0.84
Workplace nursery 0.82 0.68 0.45
Child care subsidies 0.76 0.64 0.42
High-commitment management
Teamworking 0.68 0.66 0.65
Functional flexibility 0.25 0.26 0.30
Quality circles 0.56 0.58 0.59
Suggestion schemes 0.54 0.53 0.53
Team briefing 0.79 0.77 0.76
Induction 0.67 0.67 0.68
Training in human resources 0.69 0.63 0.61
Information disclosure 0.76 0.71 0.70
Appraisal 0.52 0.50
Equal-opportunity policies
Records on ethnic origin of employees 0.90 0.89 0.90 0.88
Statistics on posts held by men and women 0.96 0.94 0.95 0.95
Promotions monitored by gender, ethnicity, etc. 0.97 0.95 0.96 0.95
Review of selection to identify indirect discrimination 0.93 0.92 0.93 0.92
Review of relative pay of different groups 0.81 0.76 0.80 0.77
Adjustments to accommodate disabled employees 0.87 0.89 0.88 0.89
No. of actual response patterns 90 90 33 1,212 61 1,907 1,311 568
No. of [(O - E )2
/E ] > 4 8 8 0 131 2 465 54 43
Maximum [(O - E )2
/E ] 8.07 8.31 0.57 25.84 5.41 32.4 16.34 81.08
% G2
explained 60.29 60.28 68.96 31.92 87.83 27.02 39.82 59.94
N 2,191 2,191 2,191 2,191 2,191 2,191 2,191 2,191
NOTE:  is the standardized alphas or discrimination parameters of the practice.
Family-Friendly Management in Great Britain / 239
hypothesis because our results show that employers approach flexible working
arrangements and child care in distinct ways. Family-friendly management
is not a single unidimensional phenomenon.
Table 5 displays the standardized coefficients (analogous to factor loadings
in factor analysis) for the latent trait model of the flexible practices.
They show that all practices contribute strongly to the latent variable, offering
the opportunity to work part time and job sharing are the strongest
indicators of this measure.
Testing hypothesis 10: The latent variable analysis of family-friendly and
high-commitment practices. Family-friendly and high-commitment practices
do not form a unity in the data. Leaving out the two family-friendly
practices that are not indicators of the underlying family-friendly management
orientation (workplace nursery and child care subsidy), neither a onefactor
latent trait model (see Table 5, model 4) nor a latent class model
fitted well. Furthermore, they do not imply two distinct factors because a
two-factor model also did not fit the data. As a consequence, we reject
hypothesis 10.
Testing hypothesis 12: The latent variable analysis of family-friendly and
equal-opportunity practices. Table 5 (model 5) shows the results of fitting
one single latent trait model to both family-friendly and equal-opportunity
practices. We again left out workplace nursery and child care subsidy. The
lack of fit of this model suggests that family-friendly and equal-opportunity
practices do not form a unity. We therefore can reject hypothesis 12.
Testing hypothesis 13: The latent variable analysis of family-friendly,
equal-opportunity, and high-commitment practices. A latent trait model of
all family-friendly, high-commitment, and equal-opportunity practice was
unsuccessful (see Table 5, model 6), suggesting that they are different underlying
concepts. This negative result could be due to the difficulty in fitting
any latent trait model to such a large number of practices (20). Yet we also
know that two of the three possible combinations of two subsets are not
reducible to a single dimension--family-friendly plus equal-opportunity
practices and family-friendly plus high-commitment practices. The analysis
of the remaining combination--high-commitment and equal-opportunity
practices--confirmed the discreteness of the three subsets of variables
(Table 5, model 7). Consequently, we reject hypothesis 13.
So far our results reject hypotheses 1, 10, 12, and 13. However, we were
able to fit a latent trait variable to the set of flexible working practices that
excluded nursery and child care subsidies practices. This latent dimension
240 / STEPHEN J. WOOD ET AL.
reflects the dependencies among only a subset of family-friendly practices.
Accordingly, we cannot treat it as a measure of a holistic friendly-family
management. We therefore will treat the latent variable based on the flexible
practices as a measure of family-oriented flexible management. We thus
have to assess the determinants of this underlying factor separately from the
determinants of the provision of child care.
In the analyses that follow, family-oriented flexible management is measured
by a latent score of each workplace in the latent dimension (respective
value of the latent variable) that resulted from fitting the logit-probit latent
trait model to the data. As in factor analysis, factor scores rank individuals
according to the discriminatory importance (as measured by the standardised
coefficients) of the practice in the model. These scores are more appropriate
than an index that aggregates the family-friendly practices. The latter
would imply that management does not discriminate between practices, an
assumption that is not supported by our results.
In order to measure child care provision, we created an ordered variable
that takes the following values: 0 if neither a workplace nursery nor child
care subsidy is provided, 1 if one or other of them is offered, and 2 when
both practices are available.
The family-oriented flexible management and child care measures are only
weakly correlated (weighted r = 0.25). This relationship is strongly affected
by the fact that 25.9 percent of workplaces in the WERS98 sample (equivalent
to 37.64 percent in Great Britain) have none of the flexible practices
or the child care practices. All those with no child care practices score below
average on the family-oriented flexible management scale. In contract, those
who have some form of child care have a 95 percent chance of being in the
top half of the distribution on family-oriented flexible management.
The family-oriented flexible management is barely correlated with the
measure of the employer's commitment to help employees balance work
and family responsibilities (r = 0.12). This implies that the systematic provision
of flexible work practices is an expression of an underlying managerial
orientation that is not strongly rooted in values. The child care measure
is more strongly correlated with this employer's commitment (r = 0.30).
Regression Analysis of the Determinants of
Family-Friendly Management
The second stage of this study aims at explaining the variability in the
provision of the family-friendly approach. Having found that flexible and
child care types of family-friendly management are discrete, we test the
Family-Friendly Management in Great Britain / 241
hypotheses about the predictors of family-friendly management on both.
We used regression analysis, employing the measures of family-oriented
flexible management and child care provision that were outlined earlier as
the dependent variables.
Further investigation of the relationships among the nine high-commitment
practices revealed that neither a one-factor nor a two-factor latent trait
model fitted the data as well a three latent class model. This implies that the
underlying orientation toward high-commitment management is categorical
rather than continuous. There is a progression along the high-commitment
scale that is not an inevitable characteristic of latent class models with good
fits. The three classes can be taken to represent minimal, partial, and full
high-commitment orientation (Table 6) and were used in the analysis as an
independent variable.
Analysis of the six equal-opportunity practices also showed that a onefactor
model could be fitted (Table 5, model 8). While two pairs of items
had residual values over four, these involved only two separate pairings, and
the percentage of G2
explained was relatively high (87.93). The latent scores
derived from this latent trait analysis of equal-opportunity practices were
used for measuring an equal-opportunity orientation.
Testing hypotheses 2 through 9, 11, 14, and 15: Predictors of the flexible
family-friendly orientation. The results of the regression analysis on
TABLE 6
LATENT CLASS ANALYSIS OF HIGH-COMMITMENT PRACTICES: THE PROBABILITY OF
HAVING A PRACTICE IF WORKPLACE IN CLASS
Model Three Classes
Class/Practice Minimal HIM Partial HIM Full HIM
Teamworking 0.39 0.64 0.90
Functional flexibility 0.32 0.40 0.70
Quality circles 0.10 0.44 0.75
Suggestion scheme 0.09 0.35 0.59
Team briefing 0.51 0.92 0.99
Induction 0.51 0.85 0.96
Training in human relations 0.09 0.45 0.79
Information disclosure 0.47 0.89 0.97
Appraisal 0.25 0.48 0.82
Percentage (weighted) 26.3 49.4 24.2
Chi-square 453.9
Log likelihood 395.1
Degrees of freedom 265
N 2,060
242 / STEPHEN J. WOOD ET AL.
family-friendly flexible management are presented in Table 7.4
They first
reveal that some of the predictors related to the institutional perspective are
significant. The adoption a family-friendly flexible management is related to
being in the public sector and health industry. It also appears to decrease
when the overall organization to which the workplace belongs is small.
However, its extent does not increase consistently with size. By contrast, the
number of employees in the workplace and whether it belongs to a larger
organization do not seem to relate to family-oriented flexible management.
In addition to the health sector, workplaces in financial services are more
likely to adopt this style than the reference category (wholesale and retailing).
As for the predictors associated with the organizational adaptation
perspective, employers' concern for the employees' work-family balance is
not significant. Only the presence of a personnel department among the
employer awareness variables is significant. Having a consultative approach
and union representation has no strong impact. When the significant variable,
personnel department, is omitted from the equation, the other measures
of a consultative approach remain insignificant. The proportion of
women in the workplace is strongly linked to family-oriented flexible management.
In contrast, having an educated workforce or a high proportion
of professional and technical employees or managers is not. There is also
no support for the business strategy thesis, since the adoption of TQM
strategy is not associated with the family-oriented flexible management.
Size of either the workplace or the larger organization does not moderate
the impact of concern on family-oriented flexible management.
None of the variables relating to the high-commitment perspective are
significantly related to family-friendly orientation. There is strong support
for the equal-opportunity perspective: An equal-opportunity orientation
has a strong positive impact on family-oriented flexible management.
Finally, only the proportion of females emerges as significant within the
situational factors perspective. The difficulty in filling vacancies and the
proportion of employees with children of various ages appear to exert no
influence on adoption of family-oriented flexible management.
The analysis thus supports hypotheses 5 and 12 concerning the proportion
of females in the workplace and equal opportunity. Key elements
of hypothesis 2 are supported, including the association of high levels of
family-oriented flexible management with the health sector. Since only one
4
Because of the high frequency of workplaces at the low end of the distribution of family-oriented
flexible management, we would have used a tobit regression. However, this model is not available in
conjunction with the STATA command that is required to handle a weighted analysis in STATA. Hence
we ran an interval regression. The results were similar to the ordinary least squares model, with the
exception of a few higher coefficients.
Family-Friendly Management in Great Britain / 243
TABLE 7
PREDICTORS OF FAMILY-ORIENTED FLEXIBLE MANAGEMENT
(WEIGHTED ORDINARY LEAST SQUARES MODEL)
 t
Institutional perspective
Employment level 0.00 0.3
Size of larger organization < 100 employees -0.35*** 2.8
Size of larger organization 100 to 999 employees -0.15 1.4
Size of larger organization 1000 to 4999 employees -0.34*** 3.5
Size of larger organization 5000 to 49,999 employees (reference category)
Size of larger organization > 50,000 employees 0.21** 2.1
Public sector 0.24** 2.1
Industry
Manufacturing -0.06 0.7
Electricity, gas and water 0.10 0.5
Construction -0.01 0.1
Wholesale and retailing (reference category)
Hotels and restaurants -0.03 0.3
Transport and communication 0.15 1.2
Financial Services 0.28* 1.7
Other business services -0.00 0.0
Public Administration 0.29 1.4
Education 0.17 1.1
Health 0.22* 1.8
Other community services -0.01 0.0
Organizational adaptation perspective
Concern for family-work balance -0.03 0.2
Concern for family-work balance × employment level 0.00 0.2
Consultative approach 0.02 0.4
Personnel department 0.29*** 3.7
Recognized union 0.02 0.3
Educated workforce -0.02 0.2
Percentage of female employees 0.51*** 3.8
Proportion of technical and professional employees 0.00 0.2
Proportion of managerial employees 0.01 1.4
Quality strategy (total quality management) 0.03 0.8
High-commitment perspective
Importance given to commitment -0.04 1.5
Importance given to human resources 0.06 1.7
Minimal high-commitment management -0.07 1.0
Partial high-commitment management (reference category)
Full high-commitment management -0.03 0.4
Enriched Jobs -0.01 0.5
Equal-opportunity perspective
Equal-opportunity practice 0.2*** 5.6
Situational perspective
High proportion of parents with children at preschool age 0.0 0.6
High proportion of parents with children under 18 -0.03 0.4
Difficulties in filling vacancies 0.01 0.3
R2
0.49
F statistics (42, 1214) 25.97***
n 1325
NOTE: Controlling for being part of a larger organization, employment change, proportion of part-timers, and
additional occupational composition measures.
*, **Significant at, respectively, 5 and 1 percent level in a one-tailed test.
244 / STEPHEN J. WOOD ET AL.
of the variables used to measure employers' awareness is significant, hypothesis
4 is only supported to a very limited extent.
Hypotheses 3 and 7--two strong hypotheses of the organizational
adaptation approach--concerning employers' values and their interaction
with workplace size are not confirmed. In addition, we tested whether the
significant employee awareness variable (personnel department) or equalopportunity
orientation mediates the impact of concern for work and
family balance, but this is not the case.
Child care provision. In order to assess the determinants of child care
provision, we conducted an ordered-probit analysis on the measure of child
care provision. The results are presented in Table 8. Only a very limited
number of variables that were significant predictors of family-oriented flexible
management have an impact on child care provision. In particular, the
presence of a personnel department, being in health and financial services,
and the percentage of women exert a positive impact on both child care
provision and family-oriented flexible management.
There are several important differences in the predictors. First, child care
provision is associated with the workplace size measure, not the organizational
one. Second, it is related to a larger range of industries. Only manufacturing
and transport and communications are less likely to adopt child
care provisions relative to wholesale and retailing. Third, the values of managers
toward the family-work balance and TQM are both significant.
Fourth, full high-commitment and two of the situational measures--
parents with children under 18 and difficulties in filling vacancies--are significantly
negatively, not positively, related to the provision of child care. In
addition to organizational size, being in the public sector and the proportion
of females have no effect on child care provision.
Conclusions
The most significant finding of this research is that in Great Britain familyfriendly
management is not an integrated phenomenon. First, our latent
trait analysis shows evidence of a systematic (unidimensional) pattern
among only one type of family-friendly practice, that connected with flexible
working arrangements. This is indicative of what we have termed a
family-oriented flexible management. Second, this form of management is
not manifested in--and hence is distinct from--the provision of child care
help. Third, it is not part of a broader management that incorporates highcommitment
management and equal opportunities. Finally, the presence of
Family-Friendly Management in Great Britain / 245
TABLE 8
PREDICTORS OF CHILD CARE PROVISION (WEIGHTED ORDERED PROBIT MODEL)
 t
Institutional perspective
Employment level 5*10-4
*1.9
Size of larger organization < 100 employees 0.25 0.3
Size of larger organization 100 to 999 employees 0.44 0.8
Size of larger organization 1000 to 4999 employees 0.07 0.2
Size of larger organization 5000 to 49,999 employees (ref. cat.)
Size of larger organization > 50,000 employees 0.19 0.4
Public sector 0.71 1.1
Industry
Manufacturing 1.8 1.6
Electricity, gas and water 5.4*** 4.7
Construction 2.7* 1.9
Wholesale and retailing (reference category)
Hotels and restaurants 2.1** 2.0
Transport and communication 1.7 1.2
Financial Services 2.5** 2.3
Other business services 3.0*** 2.8
Public Administration 2.7** 2.2
Education 3.9*** 3.4
Health 3.5*** 3.2
Other community services 3.2*** 2.8
Organizational adaptation perspective
Concern for family-work balance 1.81** 2.5
Concern for family-work balance × employment level -0.01 1.1
Consultative approach 0.09 0.4
Personnel department 1.1*** 2.8
Recognized union 0.54 1.3
Educated workforce -0.68 1.2
Percentage of female employees 3.03*** 3.6
Proportion of technical and professional employees 0.01 0.8
Proportion of managerial employees -0.02 1.0
Quality strategy (total quality management) 0.85*** 3.2
High-commitment perspective
Importance given to commitment 0.02 0.1
Importance given to human resources 0.02 0.1
Minimal high-commitment management -0.27 0.6
Partial high-commitment management (reference category)
Full high-commitment management -1.21*** 2.9
Enriched jobs 0.01 0.1
Equal-opportunity perspective
Equal-opportunity practice 0.21 0.9
Situational perspective
High proportion of parents with children at preschool age 0.14 0.3
High proportion of parents with children under 18 -1.05** 2.1
Difficulties in filling vacancies -0.70** 2.2
F (44, 1209) 3.87***
n 1322
NOTE: Controlling for being part of a larger organization, employment change, proportion of part-timers, and
additional occupational composition measures.
*, **Significant at, respectively, 5 and 1 percent level in a one-tailed test.
246 / STEPHEN J. WOOD ET AL.
family-oriented flexible management reveals nothing about management's
valuing employees' striking a good work-family balance, suggesting that the
development of family-friendly management is not underpinned by strong
normative assumptions about the family.
Management in the average British workplace is not practicing familyfriendly
flexible management. Those managements that do are likely to be in
large organizations that have personnel departments, an equal-opportunity
approach, or a high proportion of female, well-educated employee managers
among their workforces. Such managements are also more likely to provide
child care assistance, although child care is more likely to be found in large
workplaces (not necessarily organizations) and where management does
value a healthy family-work balance and have a quality strategy (as proxied
by TQM).
The extreme versions of all theories are clearly rejected by the data. The
multidimensional nature of the response of employers to an assumed societal
pressure and the lack of fit between family-friendly, high-commitment,
and equal-opportunity practices refute the institutional, high-commitment,
and equal-opportunity perspectives. The systematic pattern in the provision
of the family-oriented flexible practices questions the situational perspective.
Moreover, situational factors other than the female share of the
workforce are not important predictors of either family-friendly flexible
management or child care provision.
The organizational adaptation perspective is in many ways an umbrella
perspective and probably fares best of all the theories. Its value in pointing
to the need to empirically test the nature of the association among familyoriented
practices certainly has been vindicated. Managements do not
appear to be treating the societal pressures deriving from the changing
nature of the workforce and the family as a whole. The pressure to create
flexible working arrangements is discrete from that to create substitutes for
the time and energy of employees. The low provision of child care facilities
suggests that the former pressure is more telling than the latter (although
it is not necessarily a reason for its distinctiveness). Yet the hypotheses
concerning the determinants of family-friendly management of the organizational
adaptation perspective are supported no more strongly than the
others in the case of the family-oriented flexible management. They perform
slightly better in the case of child care provision. The proportion of female
employees, significant in both cases, suggests that local factors have some
importance.
Strong versions of the five theories analyzed here are clearly competitors,
but weaker versions can be seen as complementary. Given the lack of support
for any strong version of the theories and the heterogeneous nature of the
Family-Friendly Management in Great Britain / 247
determinants of family-friendly flexible management, we may conclude that
in Great Britain these theories are complementary. This might suggest that
attention should focus on the development of a new theory based on the
overlapping elements. Our judgment is that the organizational adaptation
perspective copes sufficiently with these overlaps to justify further attention.
The implication of our findings for future research is, however, more
fundamental. If we are to remain within an institutional perspective, we
need a theory that can explain how societal pressures are placed into specific
compartments in the process of interpretation. It may be that in certain of
these "compartments," it becomes more legitimate for employers to evoke
economic calculations in determining their response to the pressure. Paradoxically,
it may then be in these areas that employers' values become more
salient because the economic benefits are not apparent or less easily assessed.
If we apply the preceding conjecture to this study, we might hypothesize
that employers apply a cost-benefit evaluation to child care provision and
that there being no tangible net gains, a value commitment has to override
any economic assessment. In contrast, flexibility in employees' working
time and space becomes more symbolic of a meaningful response to the
societal pressure to balance family and work. The WERS98 data set lacks
data on whether employers' approach the work-family nexus in cost-benefit
terms and their perceptions of the costs and benefits of particular practices.
We are thus not able to test this conjecture.
In contrast to our results for Great Britain, analysis of U.S. data, albeit
for private-sector organizations only, suggests that family-friendly management
is more developed across the Atlantic (Wood 1999a). A latent trait
model could be fitted to a range of family-friendly practices that extended
beyond child care. While family-friendly management was associated with
similar variables to those in this study, it was also linked positively to management's
having a commitment to helping employees' work-family balance,
management's adopting a consultative approach, union representation, and
the education level of the workforce. Both company-level and workplace
measures of employment were related to family-friendly management. The
perceived positive impact of family-friendly practices was measured in the
U.S. study, and it also was significantly related to it. The proportion of
females in the workforce was not significantly related to family-friendly
management, although it was when non­child care practices were excluded
from the measure, which is the more direct comparison with the present
study. Finally, the interaction between size and management's having a
commitment to helping employees' work-family balance was significant,
suggesting that the employees' values moderate the effects of institutional
pressures.
248 / STEPHEN J. WOOD ET AL.
In short, there is much stronger support for the organizational adaptation
perspective in the comparable U.S. study. Family-friendly management
seems to be an altogether more integrated affair than it is in Great Britain.
It would appear that the majority of organizations in Great Britain are
either prior to or at what Friedman and Johnson (1996) call the stage 1 of
the work-family agenda, namely, where the issues are largely defined in
terms of mothers and child care.
The evidence from the U.S. study, however, is that family-friendly management
was still discrete from high-commitment management. This we
might expect if there were a move toward Friedman and Johnson's stages 3
and 4. The U.S. study did not include data on equal-opportunity practices.
The present study, by incorporating equal-opportunity practices into the
analysis of family-friendly practices, has revealed that a relationship
between these exists in Great Britain, although they remain discrete.
This research has gone some way toward dispelling the view that values
are vital for the development of family-friendly practices, since other factors
have been found to be important. That the proportion of females is important
for the development of both a family-oriented flexible management
orientation and the provision of child care support suggests that employers
may be partly responding to pressures and that their values are not allimportant.
The research certainly implies that the equal-opportunity
agenda can increase the extent of family-friendly management.
The lack of a significant link between high-commitment management
and family-friendly practice need not imply, given the current level of institutionalization
of the latter, that intensive employee involvement will have
no long-term effect on management's approach to the family. This is particularly
likely to be so if emphasising involvement were part of a more
integrated societal-wide initiative. A number of bodies in Great Britain
now recognize the importance of high involvement (e.g., in Great Britain
through the government's Investors in People initiative or the policies of the
Chartered Institute of Personnel and Development, as well as statutes).
Coordinating the various government programs aimed at reform in this
area and focusing them more on work-family and equality issues may yet
have an added payoff.
REFERENCES
Airey, C., J. Hales, R. Hamilton, A. McKernan, and S. Purdon. 1999. The Workplace Employee Relations
Survey, 1998 (WERS): 1998 Technical Report (Cross Section and Panel Samples). London:
Social and Community Planning Research.
Appelbaum, E., and R. Batt. 1994. The New American Workplace: Transforming Work Systems in the
United States. Ithaca, NY: Cornell University Press.
Family-Friendly Management in Great Britain / 249
Bailyn, L. 1993. Breaking the Mold. New York: The Free Press.
Bartholomew, D. and M. Knott. 1999. Latent Variable Models and Factor Analysis. New York: Oxford
University Press.
------, L. de Menezes, and T. Tzamourani. 1997. "Latent Trait and Latent Class Models Applied to
Survey Data." In Latent Class and Latent Trait Models in the Social Sciences, edited by J. Rost
and R. Langeheim, pp. 219­32. Müenster: Waxmann Verlag.
Bond, J., E. Galinsky, and J. Swanberg. 1998. The 1997 National Study of the Changing Workforce. New
York: Families & Work Institute.
Capelli, P., L. Bassi, H. Katz, D. Knoke, P. Osterman, and M. Useem. 1997. Change at Work. New York:
Oxford University Press.
Daft, R. I., and K. E. Weick. 1984. "Toward a Model of Organization as Interpretation Systems."
Academy of Management Review 9(April):284­95.
De Menezes, L., and D. Bartholomew. 1996. "New Developments in Latent Structure Analysis Applied
to Social Attitudes." Journal of Royal Statistical Society A 159:213­24.
Friedman, D., and A. Johnson. 1996. Moving from Programs to Culture Change: The Next State for the
Corporate Work-Family Agenda. New York: Families & Work Institute.
Gonyea, J., and B. K. Googins. 1996. "The Restructuring of Work and Family in the United States: A
New Challenge for American Corporations." In The Work Family Challenge, edited by S. Lewis
and J. Lewis, pp. 63­78. Thousand Oaks, CA: Sage.
Goodstein, J. 1994. "Institutional Pressures and Strategic Responsiveness: Employer Involvement in
Work-Family Issues." Academy of Management Journal 37(April):350­82.
------. 1995. "Employer Involvement in Eldercare: An Organizational Adaptation Perspective."
Academy of Management Journal 38(December):1657­71.
Hall, D. T. 1990. "Promoting Work Family Balance: An Organization-Change Approach." Organizational
Dynamics 18(Spring):5­8.
Hattie, J. 1985. "Methodology Review: Assessing Unidimensionality of Tests and Items." Applied
Psychological Measurement 9(June):139­64.
Huselid, M. A. 1995. "The Impact of Human Resource Management Practices on Turnover, Productivity
and Corporate Financial Performance." Academy of Management Journal 35(August):635­
72.
Ingram, P., and T. Simons. 1995. "Institutional and Resource Dependence Determinants of Responsiveness
to Work Family Issues." Academy of Management Journal 38(October):1466­82.
Kochan, T., and P. Osterman. 1995. The Mutual Gains Enterprise. Boston: Harvard Business School
Press.
Lawler, E. 1986. High Involvement Management. San Francisco: Jossey-Bass.
Lewis, S. 1997. " `Family Friendly' Employment Policies: A Route to Changing Organizational Culture
or Playing About at the Margins?" Gender, Work and Organization 4(January):13­23.
MacDuffie, J. P. 1995. "Human Resource Bundles and Manufacturing Performance: Organizational
Logic and Flexible Production System in the World Auto Industry." Industrial and Labor Relations
Review 48(March):197­221.
Millward, N., M. Stevens, D. Smart, and W. R. Hawes. 1992. Workplace Industrial Relations in Transition.
Aldershot, England: Dartmouth.
Milliken, F. J., J. E. Dutton, and J. M. Beyer. 1990. "Understanding Organizational Adaptation to
Change: The Case of Work-Family Issues." Human Resource Planning 13:91­107.
Morgan, H., and F. J. Milliken. 1992. "Keys to Action: Understanding Differences in Organizations'
Responsiveness to Work-and-Family Issues." Human Resource Management 31(3):227­48.
Oliver, C. 1991. "Strategic Responses to Institutional Processes." Academy of Management Review
16(January):145­79.
Osterman, P. 1995. "Work/Family Programs and the Employment Relationship." Administrative Science
Quarterly 40(December):681­700.
Porter, M. 1985. The Competitive Advantage. New York: The Free Press.
Walton, R. 1985. "From Control to Commitment in the Workplace." Harvard Business Review 63(2):77­
84.
250 / STEPHEN J. WOOD ET AL.
Wood, S. 1996. "How Different Are Human Resource Practices in Japanese Transplants in the UK?"
Industrial Relations 35(October):511­25.
------. 1999a. "Family-Friendly Management: Testing the Various Perspectives." National Institute
Economic Review 168(April):99­116.
------. 1999b. "Human Resource Management and Performance." International Journal of Management
Review 1(December):367­413.
------. 1999c. "Getting the Measure of the Transformed Organization." British Journal of Industrial
Relations 37(September):391­418.
------ and M. Albanese. 1995. "Can You Speak of a High Commitment Management on the Shop
Floor?" Journal of Management Studies 32(March):215­47.
------ and L. de Menezes. 1998. "High Commitment Management in the UK: Evidence From the
Workplace Industrial Relations Survey and Employers' Manpower and Skills Practices Survey."
Human Relations 51(April):485­515.