Soc Indic Res (2009) 93:445–467

DOI 10.1007/s11205-008-9435-0

Work-Family Conflict and Working Conditions

in Western Europe

Duncan Gallie Æ Helen Russell

Accepted: 21 December 2008 / Published online: 4 March 2009

Springer Science+Business Media B.V. 2009

Abstract This article explores the influence of working conditions on work-family

conflict (WFC) among married/cohabiting employees across seven European countries.
Using data from the European Social Survey, the paper first investigates the role of
working conditions relative to household level characteristics in mediating work-family
conflict at the individual level. It then considers whether perceived conflict is lower in
countries with coordinated production regimes and where social policy is more supportive
of combining paid work and care demands. For men the lowest rates of WFC occurred in
Denmark, Sweden and Norway, so for men there was a distinct ‘Nordic’ effect consistent
with the welfare and production regime expectations. For women, we find paradoxically
that ‘raw’ levels of work-family conflict are particularly high in France, Denmark and
Sweden where supports for reconciling work and family life are high. Our models show
that the high conflict among French women can be explained by household composition
factors and so is due to higher levels of family pressures. Higher levels of conflict among
Danish and Swedish women appear to be associated with their longer hours of work. Work
conditions are found to play a larger role than family characteristics in accounting for
work-family conflict, both in the country level models and in the pooled models. While this
partly reflects our focus on the spillover of work into family life, it is notable that family
characteristics have little effect in mediating work pressures. The results suggest that a
policy emphasis on improving work conditions is likely to have major leverage in reducing
work-family conflict.

Keywords Family and work
Work-life balance
Work conditions

European Social Survey
Job pressure

D. Gallie (&)
Nuffield College, New Road, Oxford OX1 1NF, UK
e-mail: [email protected]

H. Russell
The Economic and Social Research Institute, Whitaker Square, Sir John Rogerson’s Quay, Dublin 2,
Ireland
e-mail: [email protected]

123
446 D. Gallie, H. Russell

1 Introduction

In the course of the 1990s, work-life balance became a central issue for research into the
quality of working life. In part this derived from a common perception that greater
international competition was intensifying work and thereby reducing the time and energy
available for family and leisure life. In part, it reflected the rise in female labour market
participation and the concern that women’s integration into employment was at the cost of
assuming a double burden in which the time spent in paid work was not compensated for,
or only very inadequately compensated for, by a reduction in their domestic work. These
two potential long-term sources of increased tension between employment and private life
gave rise to rather different perspectives on the proximate source of work-life conflict—the
first emphasizing the nature of work conditions and the second the social organisation of
the household. Much of the discussion and research that flowed from this at least implicitly
assumed that the processes affecting work-family balance were similar across advanced
societies. However, a growing literature on differences between capitalist societies posed
the question of whether the relationship between the pressures of employment and the
household was likely to be mediated either by the nature of a country’s production regime
or by its welfare institutions.
Taking first the pressures from work, there has been a growing consensus that there has
been a long-term increase in work intensity. This had long been a tenet of neo-marxist
analyses of capitalist economic development. But even researchers with a more optimistic
perspective on the development of skills recognised that the demands of jobs could well
increase as a result of additional responsibilities and decision making pressure. Certainly
there is some empirical support for the view that work intensity increased in many Western
countries in the 1990s (Askenazy et al. 2006; Green 1999; Green and McIntosh 2001).
However, a more recent development has been to question whether the impact of work
pressures can be understood without taking account of the structure of households. A par-
ticularly important contribution to this discussion came from Jacobs and Gerson (2001), who
argued that it was not work pressure in the form of the length of individual working time that
was increasing the problems of work-family balance, but rather shifts in the nature of the
household—in particular the rise of dual-earner households and of lone parent households.
This suggests that it is the combination of the employment schedules of household members
rather than individual work patterns that should be the central focus of attention.
But the work time of the household is not just the aggregated time of its members in
paid employment. It also includes time devoted to domestic work and its distribution
through the division of domestic labour. A similar pressure arising from overall time in
employment could have rather different consequences depending on the demand for
domestic work. Moreover, the implications of paid employment for the pressures on
individual partners in the household are likely to be very different depending on the
domestic division of labour. This issue is highlighted in the ongoing debate on whether
rising female employment has led to a double burden for women (Gershuny 2000; White
et al. 2003; Hochschild 1990).
While arguments about the impact of work pressure and the changing household
structure tended to emphasise general determinants of work-life pressure, a growing lit-
erature on institutional variations in advanced capitalist societies has pointed to ways in
which such effects might be mediated by the forms of production and welfare regimes in
specific societies.
The production regime literature relates most directly to the likely intensity of work
pressures. Its proponents have argued that cross-national variation in production systems

123
Work-Family Conflict and Working Conditions 447

have led to quite different employer strategies for competitive advantage. In particular, a
contrast is drawn between coordinated market economies (such as the Nordic countries and
Germany) that seek to improve competitive position through an upgrading of workforce
skills, while preserving good quality employment conditions and a high level of employee
workplace control, and liberal market economies (such as the UK) that combine an
expansion of higher level skills with a low skilled workforce that is subject to an inten-
sification of work through tighter managerial control over the work process, increased
work pressure and a drive to weaken union influence, thereby increasing employee inse-
curity (see Gallie 2007a, b for reviews of this literature). Hence, the initial expectation
from the arguments of production regime theory would be that work pressures would be
greatest in countries that are closer to the model of liberal market economies.
In contrast, the ‘welfare regime’ and gender regime literatures have emphasised factors
that are likely to affect the strength of household pressures—in particular the diversity of
the levels of institutional support for childcare. In particular the Nordic societies stood out
in terms of their more generous leave provisions and their stronger infrastructure for early
childcare (Esping-Andersen 1990). In contrast, welfare systems of the Continental type
(such as Germany) sought to resolve the potential conflict between work and family life by
providing strong incentives for women to leave the labour market altogether when there
were young children in the household. Liberal regimes provided the least support for
childcare, since they offered neither good public provision nor adequate support for home-
caring.
Our paper takes advantage of new data to assess these different arguments about the
determinants of work-family conflict. We examine the relative importance of work and
family determinants and whether there are significant differences between societies, in the
light of differences in their production regimes and their policy supports for reconciling
work and family life. The paper takes the following structure: in the next section we
discuss the selection of countries and present our data; then in section three we formulate a
series of more specific hypotheses about work, family and country effects. In Sect. 4 we
present the results of our analyses and draw conclusions in Sect. 5.

2 Data and Choice of Countries

The relative importance of pressures deriving from systems of work organisation and the
organisation of domestic work and the extent to which these are mediated by differences in
broader institutional structure has received relatively little empirical investigation. In good
part, this reflects the absence of datasets with a sufficiently broad range of information
about different aspects of people’s lives. This is particularly the case with respect to
comparative data, where there has been a notable dearth even of robust comparative
measures of work-family conflict.
The creation of a special work family module in the second (2004) wave of the
European Social Survey was specifically designed to remedy this and to make possible for
the first time a rigorous analysis of the relative effects of work pressures and a range of
household constraints, in countries with diverse forms of welfare provision. The European
Social Survey provides high quality comparative data, from interviews with randomly
drawn samples of the population aged 15 and over. It uses rigorous translation protocols
and face to face interviews conducted in respondent’s homes.
To assess arguments about the possible mediating effect of country institutions we focus
on seven countries selected as relatively clear representatives of different regime types.

123
448 D. Gallie, H. Russell

We select three Nordic societies (Denmark, Sweden and Norway) as strong examples of
both coordinated market economies and dual breadwinner work/care regime with very well
developed supports for reconciling work and family. We compare the levels of work-family
conflict in these countries with the Netherlands which has a co-ordinated production regime
but a male breadwinner work/care regime and the UK which is taken as a liberal market
economy and a liberal welfare regime, with comparatively low family/work supports.
Finally they are compared with Germany and France. Germany is seen as an exemplar of a
coordinated market production regime and as a continental-corporatist welfare regime.
France is perhaps the least easily classified. It is usually regarded as an example of a
continental-corporatist welfare regime but it has a relatively high level of full-time female
employment and strong early school provision. The response rates for these countries were:
65% in Sweden; 66% in Norway; 64% in Denmark; 64% in the Netherlands; 51% in
Germany; 51% in the UK; and 44% in France. Throughout the analyses the data have been
weighted by national weights that correct for response biases within countries.

3 Work Quality, Family Structure and the Experience of Work Family Conflict:
Hypotheses

In a context of rapid technological change and increased international competition,

European employers have been under greater pressure to improve the efficiency of their
production processes. This has led to an increased emphasis on raising the skills of the
workforce, intensifying the work process and increasing flexibility in the dismissal of
employees. There are grounds for thinking that skill and responsibility levels, work
pressure, and insecurity are major factors affecting work stress and are therefore likely to
have important on-flow effects for pressures on family life. We consider these briefly in
turn.

3.1 Skill and Responsibility

There has been a long term rise in the skill profile of the workforce in most capitalist
societies, reflected in part in a growth in the number of employees in managerial and
professional occupations (Tahlin 2007). In general those at higher occupational levels will
have more complex tasks with higher levels of responsibility, which could lead to work
preoccupations spilling over into family life. Although this may be partly offset by the fact
that such jobs provide a greater sense of control, self-confidence and organisational skill
that may mediate the impact of such pressures, employees in these jobs are also likely to
work longer hours. The combination of increased responsibility with longer work hours
can be expected to lead to relatively high levels of work-family conflict.
At the same time, it has been suggested that the development of new forms of work
organisation is associated with an increased delegation of responsibility to employees,
leaving them with greater task discretion or initiative to take decisions about the job. This
could be expected to lower work pressure and hence the risk of work family conflict
(Karesek and Theorell 1990).
Our main hypotheses with respect to skill are then:
1. Higher skilled employees will tend to experience higher work-family conflict (WFC)
than lower skilled employees.
2. Higher levels of control may reduce WFC and so offset some of the high skill effect.

123
Work-Family Conflict and Working Conditions 449

3.2 Work Pressure

There are four aspects of work pressure that are likely to have implications for employee
stress: working hours, the intensity of work, the safety of work conditions and job
insecurity.
Long working hours are likely to be a major pressure on family life because they affect
the level of physical exhaustion experienced by employees and generate greater time
budget pressures for managing family activities. Employees working part-time, in contrast,
should experience lower levels of work-family conflict. The scheduling of work hours is
also likely to be related to work pressure. The growth of a service economy and the
increased use of capital intensive equipment create pressures for an extension of working
hours into times that were traditionally reserved for the family. Where employers are free
to extend working times into ‘unsocial’ work hours, one could expect particularly high
levels of work-family conflict. However, where longer working hours are combined with
flexibility they may have less serious consequences.
Work intensity has traditionally been regarded as reflecting the degree of control by first
line supervisors and pay systems that closely relate reward to effort. But it has become
increasingly linked to the introduction of targets with demanding deadlines, and in the
service sector, by customer control of work rhythms. The spread of new management
techniques, such as just in time production, and of more detailed monitoring of perfor-
mance through new technologies, may have further raised effort levels at work.
The pressures of work may be also related to the physical quality of the work envi-
ronment. Where work conditions are unsafe, employees are likely to have to work with
particularly high levels of concentration. Unsafe work conditions are also likely to be
associated with more arduous physical conditions that will increase levels of exhaustion.
Finally, job insecurity is likely to add significantly to the pressure experienced in work.
This is partly due to the strain imposed by insecurity itself. But it may also make
employees more vulnerable to work intensification.
The main hypotheses with respect to work pressure are then that:
1. Longer work hours and more extensive unsocial work hours will increase work-family
conflict.
2. A greater prevalence of part-time work will reduce work-family conflict for women.
3. Higher work intensity will increase work family conflict.
4. Less safe work conditions will heighten work family conflict.
5. Lower job security will increase work-family conflict.

3.3 Family Arrangements

Turning to the other side of the work-family equation, the arguments in the literature point
primarily to two sources of strain deriving from family relations. The first relates to the
overall pattern of time demand on the family. This may result from the employment
commitments of the different family members, with the expectation being that work-family
conflict will be highest where both partners are engaged in full-time employment (Jacobs
and Gerson 2004). But it is also likely to depend on the housework and caring demands
within the family. In particular, where families have dependent children, the pressure on
family time budgets is likely to be much higher (Folbre and Bittman 2004). The second
potential source of strain relates to the way domestic work is organised within the family.
The expectation is that where male partners respond to female involvement in paid work by

123
450 D. Gallie, H. Russell

taking over a higher share of domestic and caring work, women’s experience of work-
family conflict should be correspondingly reduced.
Family employment patterns and internal work organisation may affect work-family
conflict not only through the independent pressures they pose, but also through their
potential for reducing or accentuating the impact of work demands. There is a significant
literature pointing to the importance of social support in mediating work pressure and the
family is clearly a potentially important site of such support (Theorell 1998). Different
patterns of family organisation, by affecting the mutual availability and preoccupation of
partners, may result in the family being differentially effective in buffering the effects of
work demands. In general, it could be expected that the greater the time pressures on the
family, the less likely it is to be in a position to mediate work pressures.
The central hypotheses with respect to family effects are that:
1. Greater family commitments in the form of children, especially younger children and
high levels of household work should increase work-family tension.
2. WFC will be higher in dual earner households (measured by the variable ‘partner
employed’). Partner’s hours and work schedule are also expected to impact on WFC.
3. The ‘double burden’ model suggests that conflict will be higher for women and that lower
levels of conflict will only emerge where men take on a greater share of domestic work.
4. At any given level of work pressure, we would expect those in dual earner households
and in households with children to experience higher levels of WFC than those in less
time pressured households.

3.4 Country Hypotheses

The major source of difference between countries in work-family conflict that would be
expected on the basis of production regime theory is the level of work pressure. Employers in
liberal market economies are less constrained by the need to win commitment from highly
skilled workers, they are less subject to the institutional controls of works councils or trade
unions, and they have greater power over their workforce through their freedom to hire and
fire with relative ease. It could be expected then that they would be likely to adopt policies that
would subject employees to higher levels of work pressure. Overall, the main hypothesis
deriving from a production regime approach would be that British employees will be subject
to greater work-family conflict as a result of higher work pressure and this will be particularly
the case for male employees due to the prevalence of long working hours.
The literature on welfare regimes and policies highlights the extent to which the
opportunities to combine work and caring and the gender division of paid and unpaid work
vary across societies. Liberal welfare regime such as the UK have typically been char-
acterised as intervening less in the family than either conservative or social democratic
regimes. The UK has also been classified as a strong male breadwinner regime (Lewis
1992). Caring is seen primarily as a private endeavour, with state support for childcare
historically low compared to elsewhere in Europe, although there has been a significant
investment in childcare support in Britain in recent years (Lewis 2006). Low benefit levels
in these societies mean that support for full-time caring is poor, which places strong
pressure for women to work at least part-time. Work-family conflict could be expected to
be high in liberal regimes not only among women but also among men due to longer male
working hours.
Conservative welfare regimes are thought to promote a more traditional division of
labour between the sexes through the tax and welfare systems (e.g., Germany, Netherlands,

123
Work-Family Conflict and Working Conditions 451

France). This may make it easier to reconcile work-family demands, due to the reduced
pressure of paid work on the female partner. But the role of male breadwinner may also
lead to higher work-family conflict among men, due to longer working hours to sustain
family incomes. It is important, however, to note that there is wide variation in the policy
supports for work/family reconciliation within this set of countries. For example, France
and the Netherlands provide significant support for female employment, through childcare
services in the former and flexible and part-time working arrangements in the latter (Lewis
1992; Plantenga and Remery 2005; Strandh and Nordenmark 2006).
The Nordic countries represent societies in which the state plays a large role in the
family and labour market, there is a high degree of individualisation in the taxation and
welfare systems, a dual breadwinner model is promoted, and there are highly developed
supports for combining work and caring (Crompton et al. 2007; Gornick et al. 1997; Lewis
1992). Generous leave provision, and publicly provided childcare in particular are cited as
indicators that these countries facilitate the reconciliation of work and family life (Plant-
enga and Remery 2005). Social policy has also begun to encourage greater gender equality
in caring, for instance through paid parental leave for fathers (Bruning and Plantenga
1999).
The implications of the Nordic welfare model for work-family conflict are open to
conflicting interpretations. While dual earner families are thought to be exposed to higher
levels of work-family conflict, there is a common expectation that such pressures are offset
by policies that facilitate work-family balance (Gornick et al. 1997; Gornick and Meyers
2003; Strandh and Nordenmark 2006). Hence employment in these societies should be less
problematic for women than in liberal market societies.
However, as Scherer and Steiber (2007) have argued, a welfare system that provides
extensive childcare support may be more likely to draw into the labour market households
whose conditions for resolving work-family conflict are particularly difficult. Others have
argued that the dual breadwinner model has increased women’s integration into paid work,
without leading to a corresponding involvement of men into unpaid domestic and care
work, therefore resulting in women bearing a ‘double burden’ of labour (Lewis 1992) and
hence higher levels of WFC.
Our main hypotheses about potential country differences are then:
1. Higher levels of work pressure in the UK will lead to higher levels of work family
conflict than in countries with other types of production regime.
2. Policy supports for dual breadwinning in the Nordic countries should lead to lower
work family conflict in these countries than in market oriented regimes.
3. However, the effects of such policy supports may be offset by selection effects or the
lack of adjustment of the domestic division of labour to the enhanced integration of
women in the employment structure.
4. Levels of conflict in conservative/male breadwinner regimes should be intermediate
because reconciliation occurs through a traditional gendered division of labour. But
the male breadwinner model will produce higher levels of work-life stress among men
because of longer work hours.

4 Results: Determinants of Work-family Conflict

Our measure of work-family tension focuses on four questions that seek to tap the psy-
chological and time pressures of trying to reconcile work and family life. They asked:

123
452 D. Gallie, H. Russell

How often do you

1. … keep worrying about work problems when you are not working?
2. … feel too tired after work to enjoy the things you would like to do at home?
3. … find that your job prevents you from giving the time you want to your partner or
family?
4. … find that your partner of family gets fed up with the pressure of your job?
There are five possible response categories: never, hardly ever, sometimes, often, or always.
An additional category ‘don’t have partner/family’ was allowed for items three and
four. The alpha for the scale is .74 in the pooled sample for the seven countries and ranges
from .69 in the Netherlands to .79 in the UK.1
The first two items refer to the extent to which work intrudes on life generally, while
items three and four refer to partner and family and hence are more appropriate for those
within couples. We have therefore restricted our analysis to those living with a partner. The
measures are primarily concerned with the spill-over from work to family life rather than
the intrusion of family into work or the broader and more nebulous concept of work-life
balance (Pichler 2008). This is consistent with the main policy focus to date, which has
been on how the ‘costs’ of work have fallen on the family. Moreover work to family
interference has more salience in people’s lives (Crompton et al. 2007b). In the ESS
module we find that less than 2% of the selected respondents report that they often find it
difficult to concentrate on work because of family responsibilities compared to at least 20%
on each of the first three items we have selected and 9% on the fourth.
The WFC measure applies only to the employed, it is also an individual level measure
and should not be interpreted as the level of stress for the household as a whole. The
analysis is based on an overall sample of 4,837 employees, with country samples ranging
from a low of 577 in Denmark to 822 in Germany.
We examine first the broad pattern of relationships between the household and work
variables and the index of work family conflict, and then turn to examine the nature and
determinants of country differences.

4.1 Effects of Working Conditions

Social class (EGP) is taken as the main measure of skill level (Tahlin 2007). The net effects
on work-family conflict for the pooled sample are shown in Table 1.2 The results confirm
the hypothesis that those with lower skill levels experience less work/family conflict. As
this remains the case even with the hours of work controlled, this effect is likely to be
associated with the level of responsibility within one’s job. It is notable that the class effect
emerges powerfully in each of the countries.
Task discretion is measured by a question that asks people how much ‘the management at
your work allows you: (1) to decide how your own daily work is organised, (2) to influence
policy decisions about the activities of the organisation and (3) to choose or change your pace
of work? We have created a summary index of task discretion by taking the employee’s
average score across the three items.3 Our results show that higher task discretion also
generally increases work/family conflict, which does not support the view that autonomy in
work will offset higher work pressures or will make it easier to manage the combined

1
We use raw score versions of variables throughout.
2
Details of effects within countries are available in the Appendix, Table 7.
3
The alpha for the scale is .73 for the pooled sample and ranges from .68 in Sweden to .79 in Germany.

123
Work-Family Conflict and Working Conditions 453

Table 1 Work family conflict among married/cohabiting employees

Work conditions only Family only Full model

Female 0.20** 0.24**

Child less than 6 years 0.11** 0.09**
Child 6–17 years 0.07* 0.07**
Weekly hswk hours for HH 0.003** 0.003**
Partner Employed 0.09* -0.01
HH income contribution (%) 0.11** 0.02*
Partner \ quarter hswk w/day -0.08* -0.02
Partner \ quarter hswk w/end -0.11** -0.07*
Lower professional-managerial -0.06(*) -0.08*
Routine non-manual -0.15** -0.20**
Skilled manual -0.34** -0.29**
Non-skilled -0.34** -0.31**
Hours per week 0.01** 0.01**
Unsocial hours index 0.16** 0.16**
Task discretion index 0.01(*) 0.01
Can decide start/finish time -0.03** -0.03**
Job pressure index 0.25** 0.24**
Health/safety at risk 0.15** 0.15**
Job is secure -0.05** -0.06**
Partner paid work hours 0.00 -0.00
Partner unsocial hours 0.03* 0.03*
Constant 1.29** 0.97**
Adjusted R2 0.295 0.052 .312

Note All models include controls for country. If partner is not employed hours are set to zero
Reference categories: male, no children under 18 years, partner not employed, higher professional/mana-
gerial, permanent contract. ** P \ .005; * P \ .05; (*) P \ .10

demands of family and work. It is significantly positive in four of the countries and it is only
in Sweden that we find the expected effect of it reducing work-family conflict.
Work pressure is defined in terms of four central components: the length and scheduling
of the hours of work, the intensity of work, the safety of work conditions and job security.
Unsocial hours are captured by three questions on the frequency of weekend work, evening
work and overtime, these are combined to form a scale ranging from 1 to 5, where 1
represents those who never engage in these three activities and 5 represents participation in
all three on a weekly basis. There were three questions in the survey that bore directly on
the level of work intensity: they asked people how hard they worked, whether they
experienced acute time pressures in carrying out the job and whether their pay was directly
related to work effort. To provide a summary measure, we have created an index of work
pressure based on the average score of the three items.4 Finally there is a measure of the
safety of working conditions: ‘My health or safety is at risk because of my work’.

4
The alpha is .40 for the pooled sample and ranges from .32 in Germany to .45 in the UK, Norway, Sweden
and the Netherlands. This is low by conventional standards.

123
454 D. Gallie, H. Russell

Both the hours of work and the scheduling of these hours have a substantial influence on
the level of work family conflict. The longer the hours of work the greater the experience
of work-family conflict, and those who work in the evenings and at weekends are also
particularly likely to find it difficult to reconcile their work and family lives. This is the
case in all countries, although the effect is especially strong in the UK. Working unsocial
hours significantly increases WFC independently of the number of hours worked. A strong
relationship is again found in all countries between work intensity and work-family con-
flict. However discretion over working time reduces WFC. Unfavourable mental or
physical conditions that put employees’ health and safety at risk are associated with
heightened WFC.
Our main job security indicator is a question that asks respondents on a four item scale
how true or not it is that ‘My job is secure’. Job security has a weaker effect on WFC than
skill level and work pressure. It has, however, a clear effect in reducing work-family
conflict in the overall sample and is significant in the within country analyses for the UK,
Denmark, Norway, Germany and Netherlands.
Overall, work characteristics have a substantial effect on work-conflict scores,
accounting for some thirty percent of the overall variation for both men and women
(Table 1).

4.2 Family & Household Characteristics

The ESS data includes rich data on the level and organisation of household work along
with information on partners’ working hours schedules. Five household variables are
included in the analyses. The first is the age of children (child aged 0–5, child aged over 5).
Second we include the hours of housework undertaken by household members per week—
this reflects the demand for domestic work from the household as a whole. Third we
include the proportion of household income provided by the respondent, which indicates
the extent to which the respondent is the primary breadwinner. Fourth, we control for
partner’s employment status. Finally we include a measure of partner’s contribution to
housework. Indicators of partner’s working hours and schedules are included in the work
conditions model.
For the overall sample, all of these variables are found to be significantly associated
with WFC. Having children increases conflict, and younger children have a stronger
impact than older children. The effect of having dependent children is stronger among
women. In the country models it is significant only in Denmark, Germany and Norway.5
Increasing hours of unpaid work add to the risk of work-family conflict, due to time budget
constraints. Having a partner in employment also increases WFC as predicted by Jacobs
and Gerson (2001). However, the partner with greater financial responsibility experiences
higher WFC. Our expectation was that having a partner who did not share the domestic
workload would add to work-family conflict especially for women. The results in model 1
suggest in contrast, that where a partner does less of the housework WFC is lower.
However, share of housework is correlated with partner’s hours of paid work and when this
is controlled, partner’s contribution to housework on weekdays is insignificant. Partner’s
paid work hours do not significantly affect WFC but a partner working unsocial hours adds
to the respondent’s own sense of work-family conflict.
The net effects for the overall sample are shown in the second model of Table 1. When
family characteristics are controlled, women experience higher WFC than men. All

5
Details of these country models are available in Appendix Tables 7 and 8.

123
Work-Family Conflict and Working Conditions 455

household factors have significant net effects, further, as can be seen in the last model of
Table 1, most remain significant even when the work characteristics are taken into account.
Overall household effects account for only 5% of the variance. While the effect is
somewhat stronger for women (8%) than for men (4%), family characteristics are less
effective at explaining the spill-over from work to family life than working conditions for
both sexes. Moreover, the effects of work factors are little affected by family circum-
stances. The coefficients of work factors change little when household factors are
controlled and remain statistically highly significant. Nor is there evidence to support the
hypothesis that the effects of working conditions are moderated by family characteristics.
Neither the interactions between the work pressure indicator and presence of children and
employment status of partner nor the interactions between unsocial hours, work time
flexibility and family characteristics were significant. The interaction between hours of
work and presence of children was also insignificant. However, the effect of working hours
on WFC was found to be moderated by partner’s employment status with the detrimental
impact of hours of work being greater in dual earner households.6

4.3 Country Effects

Taking the country work-family conflict scores (Table 2), it can be seen that overall the
lowest levels of work-family conflict were in the Netherlands and Norway. The highest
level of work-family conflict was in France, followed by the UK and Germany, while
Sweden and Denmark were in an intermediary position.
However, a test of statistical significance showed that only the Netherlands, Norway and
Sweden have lower work-family conflict than the UK. The overall pattern does not fit well
then with the expectation that the UK would stand out as the country with the highest
work-family conflict.
These overall country figures, however, conceal rather different patterns by sex. Dif-
ferences between countries are notably stronger among men than among women. Work-
family conflict is significantly lower for men in the Netherlands, Denmark, Norway and
Sweden. It is particularly high for men in Britain, France and Germany. The pattern of
country differences for men suggests a distinct ‘Nordic’ effect, consistent with the
expectations derived from welfare regime theories. The ‘liberal’ and ‘conservative
regimes’ (typified respectively by the UK on the one hand and Germany and France on the
other) were expected to impose particular strain on the male breadwinner.
The predictions for women were less consistent. On the one hand, better childcare and
leave provision might be expected to reduce the problems of reconciling work and family
life in the ‘social democratic’ welfare states. On the other, the greater opportunities for
participation in the labour market and high levels of employment within couples might
increase strain. In practise, the pattern does not fit well with either expectation. While
women in Norway have lower work-family conflict, those in Denmark have higher, while
Swedish women are not significantly different from British. Clearly the ‘high welfare’
regimes of the Nordic countries do not lead necessarily to low levels of work-family
conflict for women. Furthermore, it is in France, rather than in Britain, that women
experience the most severe strain in reconciling work and family life.
We next examined whether the differences between country patterns were due to the
prevalence of specific employment practises or household characteristics. While there was

6
The coefficient for the interaction between hours of work and partner employed is .005, the main effect of
working hours remains significant and also has a coefficient of .005.

123
456 D. Gallie, H. Russell

Table 2 Work-family conflict scores by country, employees in couples

All Men Women

Germany 2.62 2.70 2.52

Denmark 2.54 2.46 2.62
France 2.66 2.65 2.68
UK 2.63 2.75 2.51
Netherlands 2.41 2.41 2.40
Norway 2.47 2.52 2.40
Sweden 2.56 2.53 2.59

Table 3 Work and family characteristics of employees in couples

DK FR DE UK NL NO SE

Work characteristics
Higher professional-mangerial (%) 24.0 19.8 21.3 20.5 26.4 22.8 20.0
Skilled manual (%) 8.4 15.7 9.9 8.5 11.1 11.6 12.5
Non-skilled (%) 17.3 13.6 12.7 17.4 10.1 12.6 14.6
Mean task discretion score (score) 6.33 5.87 5.25 5.43 5.89 6.69 6.42
Work hours 50? (%) 9.6 11.8 14.9 19.7 12.0 10.0 9.5
Evening work weekly? (%) 20.3 17.6 13.0 27.2 21.8 28.1 26.9
Overtime weekly? (%) 17.3 18.0 23.6 16.9 12.6 13.2 26.9
Works every weekend (%) 5.2 9.4 4.9 14.2 3.5 2.6 3.6
Work intensity index (score) 2.93 2.86 2.90 3.10 2.77 2.95 3.01
Very true can decide time start/finish work (%) 18.0 34.2 12.1 12.2 12.6 17.6 19.0
Family characteristics
Both employed (%) 74.2 65.8 56.7 61.0 66.1 74.7 77.2
Male breadwinner (%) 14.3 19.3 28.9 24.5 23.2 15.9 14.3
Men contribute very large % of HH income (%) 17.8 23.5 48.4 42.6 43.9 19.2 16.6
Women, children \ 18 years (%) 49.5 62.1 37.9 46.1 46.2 45.5 49.7
Weekly housework hours 32.7 23.1 32.7 30.0 30.7 29.3 32.6
Men’s share housework (%)a
None/almost none 14.2 15.5 29.2 14.4 17.3 7.8 6.5
Up to a quarter 25.0 32.1 32.1 34.6 30.8 31.0 19.4
Quarter to half 47.4 36.2 28.6 36.5 36.5 49.4 59.9
More than half 13.4 16.2 10.1 14.4 15.4 11.0 14.2
a
Men’s share of housework at weekends in dual-earner households

little systematic difference in the proportions in higher professional and managerial work,
task discretion was markedly higher in the Nordic countries (Table 3). Long working hours
and weekend work were notably more common in the UK, though it should be noted that
the pattern of working hours is highly polarised by gender. At the same time, British
employees are much less likely than those in the Nordic societies to have flexible time
arrangements. The UK also had the highest level of work intensity measured by how hard
people worked, the time pressures in their work and the extent to which work effort was
related to pay and this was particularly the case for male employees.

123
Work-Family Conflict and Working Conditions 457

With respect to family characteristics, the notable difference is the much higher pro-
portion of dual earner households in the Nordic countries. Only 63% of German male
employees have an employed wife, compared to 84% of those in Sweden. The male
contribution to household income follows the same country pattern. It is highest for
couples in Germany, the UK and the Netherlands and is lowest in Sweden, Norway and
Denmark. Table 3 also outlines the level of male involvement in household work in dual
earner households, as reported by both men and women. In Germany, 60% of respondents
report that the male partner does less than a quarter of the housework, this is also true for
around half of dual earner households in the Netherlands, France and the UK. The Nordic
countries report a higher male share of housework, with Swedish respondents reporting the
most equitable division of housework across the sexes. However even in Sweden there are
few households that report men doing half or more of the housework. There would appear
then to be some, but very imperfect, adaptation of the household division of labour to the
dual earner working pattern, suggesting that the high level of dual employment may still be
associated with work-family conflict. In Denmark, Sweden and, most strikingly, France a
higher proportion of female employees have dependent children. This is consistent with the
argument that policies in these countries lead to the inclusion in the labour market of
women with greater family demands.
To examine the impact of such differences in family and work characteristics, we
compare the initial differences in country coefficients to those when first family/household
factors are controlled and second when working conditions are taken into account
(Table 4). The results are shown for all employees and for male and female employees
separately. However, as the expectations about such effects are rather different for men and
women, we focus the discussion on the sex-specific models.
For women, controlling for family characteristics reduces the WFC coefficient for
French women as expected and it is no longer significant. The fact that France stood out as
the country where women had the highest levels of work-family conflict can be linked to
higher levels of family pressures. In Denmark, there is also sharp reduction in the coef-
ficient, which becomes non-significant. In Sweden, controlling for family factors turns a
positive coefficient into a negative one. When family characteristics are controlled, female
employees in the Netherlands and Norway stand out as having significantly lower work-
family conflict than women in the UK.
It may also be that the patterns are affected by differential selection into employment of
women confronting particularly high levels of domestic difficulty. The anticipated effects
of strong state childcare provision may be counterbalanced by the fact that it enables
women to enter employment who in other countries would find this impossible.
To examine this, in a supplementary analysis a selection equation was calculated for the
final joint country model of WFC for women. Selection into employment was not sig-
nificant in the female model.7 This suggests unobserved differences in the ability to
reconcile work and family conflict, due to national differences in female labour market
participation, do not account for the remaining country differences in the final model.
Turning to work conditions, the most striking effect is for Swedish women. When work
conditions are controlled, the score for Swedish women becomes significantly lower than
that for British women. Taking account of work conditions also strengthens the significant
negative coefficient for Norwegian women and turns the coefficient for Danish women
from positive to negative. This suggests that one factor that accounts for the initial lack of

7
The participation model included country, age, age squared, years of education and number of children.
See Appendix Table 9.

123
458 D. Gallie, H. Russell

Table 4 Work family conflict: changes in country coefficients

Model 1 Model 2 ? family Model 3 ? work
country only controls conditions

All
Denmark -.07 -.12* -.07
Norway -.16** -.19** -.14**
Sweden -.08(*) -.12* -.13*
Germany .00 -.04 .01
NL -.22** -.26** -.10*
France .04 .04 .12**
Women
Denmark .12(*) .00 -.08
Norway -.12(*) -.18* -.20**
Sweden .07 -.06 -.17*
Germany .03 -.04 -.03
NL -.11 -.14* -.05
France .13(*) .08 .08
Men
Denmark -.31** -.30** -.14*
Norway -.25** -.24** -.14**
Sweden -.25** -.24** -.16**
Germany -.04 -.05 .02
NL -.34** -.35** -.17**
France -.09 -.06 .03

Note: The UK is the reference country

** P \ .005, * P \ .05, (*) P \ .10

difference between women in the Nordic societies and those in the UK is that the former
have more demanding and stressful work conditions.
The contrast with the effects of introducing the same set of controls for men is inter-
esting. Consistent with the predictions of production regime theory, the initial model for
men, without controls, shows that men in the Netherlands and the Nordic Countries (but
not in Germany) have relatively low levels of WFC. Controlling for family characteristics
makes little difference to the country coefficients for men. However when work charac-
teristics are introduced the differences between the Nordic countries and the UK are
reduced substantially. This change suggests that superior working conditions in the Nordic
countries are responsible for an important part the difference.
In Table 5 we further disaggregate the work condition effects to test which drive the
changes in country coefficients. We test the effects on the initial country coefficients of
introducing controls for two aspects of work pressure (work hours and work intensity) on
the one hand and class composition on the other. It can be seen that class controls make
very little difference to the pattern for men. It is hours of work and work intensity that are
the strongest factors in accounting for lower work-life conflict among men in Denmark, the
Netherlands, Norway, and Sweden. In the Netherlands, the story is straightforward—male
employees have relatively short work hours and low levels of work pressure. Lower
average hours of work also play a role in accounting for the lesser prevalence of work-life
conflict among Swedish men. Work pressure scores for men are high, but the relationship

123
Work-Family Conflict and Working Conditions 459

Table 5 Changes in country coefficients work condition effects

Those in couples only

Country Hours Work Class Hours ? work

only intensity intensity

Women
Denmark .12(*) -.02 .14* .08 .03
Norway -.12(*) -.19** -.06 -.14* -.12*
Sweden .07 -.08 .08 .06 -.03
Germany .03 -.05 .13* .06 .05
NL -.11 -.04 .02 -.17* .05
France .13(*) .03 .22** .14* .14*
Men
Denmark -.31** -.23** -.21** -.33** -.17*
Norway -.25** -.19** -.19** -.26** -.15*
Sweden -.25** -.18** -.21** -.25** -.17**
Germany -.04 -.01 .01 -.05 .02
NL -.34** -.27** -.21** -.39** -.18**
France -.09 .00 -.01 -.10 .04

Note: Samples excludes employees missing on any of variables in final model in Table 2 (which includes all
work condition and family variables). The UK is the reference country
** P \ .005, * P \ . 05, (*) P \ .10

between pressure and WFC is somewhat weaker than in other countries, as is also the case
for Norway. These results support the hypothesis that in liberal market economies long
work hours lead to a particularly high level of WFC among male employees at least
compared to the North European countries.
For women hours of work is the work characteristic with the strongest effect on work-
family conflict. With hours controlled, women in Denmark and France no longer have
higher levels of conflict than those in the UK. Higher work-family conflict in these
countries was therefore partly due to women working longer hours compared to British
women (together in the case of France with differences in family factors). The fact that in
Sweden a positive coefficient becomes negative when hours are taken into account sug-
gests a similar pattern, although the coefficients are not statistically significant. Similarly,
controlling for working hours strongly accentuates the negative coefficient for Norwegian
women. In contrast, controlling for work intensity sharply reduces the difference between
Norwegian and British women, suggesting that better work conditions in the form of lower
work intensity, may help account for the fact that Norwegian female employees are rel-
atively well protected against work-family strain.8

5 Conclusions

Our point of departure was to examine contrasting accounts of the sources of work-life
conflict that have emphasised, respectively pressures from work on the one hand and

8
The mean work pressure score for female employees in Norway is 2.86 compared to 3.02 in Britain.

123
460 D. Gallie, H. Russell

pressures arising from family structure on the other. The strategy was to examine these
patterns across countries that are thought to have quite diverse institutional systems:
Denmark, Norway, Sweden, the Netherlands, Germany, France and the UK using data
from the ESS.
The analysis has confirmed that working conditions have a strong influence on the level
of work-family conflict within all seven countries. Work pressure (intensity and hours/
unsocial hours) has the most negative impact on balancing work and family demands. It
seems likely that the marked rise in levels of work pressure in European countries since the
early 1990s has contributed substantially to greater strain in managing work-family life. Our
evidence also supports the view that the rising skill trends in advanced societies, associated
with the growth of ‘knowledge-based’ economies, accentuates work-family conflict. It had
been expected that this might be partly offset by the greater responsibility and task dis-
cretion that is associated with more skilled work, but this did not prove to be the case.
Family characteristics have a much weaker influence on WFC than work. Increased
caring and housework demands heighten WFC but the proportion of variance explained by
family characteristics is much lower than in the case of job characteristics. This is partly
because the indicator of WFC used focuses on the spill-over from work to family life, and
is consistent with the findings of Byron (2005).9 However, this is the primary form of
work-family conflict that people report. It is also notable that family characteristics have
little effect in mediating the impact of work pressures The results suggest that changes in
employment practises are likely to be to the key factor in reducing work-family conflict.
Our second concern was to assess whether there were significant country variations in the
prevalence of work-life conflict, and, if so, whether these could be attributed to differences
in either production or welfare regimes. The respective theories could be seen as particularly
relevant to the strength of specific work and family pressures. Our expectations here,
following the broad logic of production regime theory, was that Britain would have an
exceptionally high level of work pressure and this should raise work-family strain. Our
expectation from welfare regime theory was that male employees in the UK and Germany
would be under greater pressure as a result of their breadwinner roles. There were alter-
native predictions of the implications of welfare systems for female employees. Some
accounts suggested that the superior public provision of caring in the Nordic societies would
provide stronger protection for women from work-family conflict than would be the case in
liberal market regimes. However, others predicted a particularly high burden on women in
these societies due to the fact that they encouraged dual earner families, in a context in
which women still remained largely responsible for domestic work within the household.
National differences point more to the distinctiveness of the Northern European soci-
eties than to differences between ‘production regimes’. However, even this effect is only
the case for male employees. Men had the lowest level of work family strain in Norway,
Sweden, the Netherlands and Denmark. A key feature of these economies is that they do
not impose long working hours on male employees and workers are given greater flexi-
bility in their working times. These positive features appear to outweigh the potentially
negative side-effects of the high skills approach and of greater responsibility. Similarly,
although work pressure is unexpectedly high in some of the Nordic societies, the
employment regimes in these countries are more successful in moderating the effects of
pressure so it does not translate so directly into WFC.

9
In this meta analysis of 61 studies, Byron found that work variables have a stronger effect on work to
family interference than family to work interference. However, family variables did not have a consistently
stronger relationship with family to work interference (2005, p183).

123
Work-Family Conflict and Working Conditions 461

In contrast, the level of work-family strain varied relatively little for women and such
variation that existed did not fit well with any particular regime expectation. The highest
level of strain was among French female employees, followed by the Danish and the
Swedes. It is clear that the welfare regimes of the Nordic countries, with their stronger
emphasis on caring provision, did not lead necessarily to lower levels of work-family
tension. There was no evidence that having children was a lesser source of pressure in these
societies.
The origins of WFC in the workplace may help explain why distinctive policy
approaches to care/parental supports do not result in the expected benefits for women
employees in Nordic countries. Women in these countries are more integrated in the
workforce, but this is associated with longer working hours and higher levels of job
pressure. The relatively long hours of work and high job pressure cancel out any impact of
greater support for caring. In contrast, in the UK and the Netherlands, family pressures
were partly absorbed by the fact that a high proportion of women worked part-time. The
distinctiveness of work-family strain among women in France may reflect the fact that it is
a society that encourages high female labour market participation, while providing neither
the level of public care support available in the Nordic societies nor the level of short hour
part-time work provision to be found in Britain and the Netherlands.
The analyses presented here have a number of limitations. The focus is on current work-
family tensions and there may well be other current and future ‘costs’ which cannot be
addressed within this paper. For instance, some households may have minimised conflict at
the cost of postponed or limited family formation. Similarly there may be longer term
labour market costs of reconciling work and family life in a particular manner, for example
with respect to wages and career development, which cannot be taken into account here.
Addressing these issues would require the availability of new types of data, in particular
new types of longitudinal data, and hence must remain for the research agenda of the
future.

Acknowledgments This paper was produced as part of the Economic Change, Quality of Life and Social
Cohesion (EQUALSOC) Network of Excellence, funded by the European Commission (DG Research) as
part of the Sixth Framework Programme. See editors’ introduction for further details. The authors are most
grateful to Frances McGinnity, Chris Whelan and Nadia Steiber for their insightful and constructive
comments on earlier drafts of this paper. We would also like to thank the other members of the EQUALSOC
workshop held in Dublin in October 2007 for their helpful contributions on the paper.

Appendix

See Tables 6, 7, 8, 9.

Table 6 Details of work condition and family indicators

Question wording Response category

How much the management at your work allows you 0 No Influence

(1) to decide how your own daily work is organised 10 I have complete control
(2) to influence policy decisions about the activities
of the organisation and
(3) to choose or change your pace of work?’.
Regardless of your basic or contracted hours, how many hours Hours
do you normally work in a week (in your main job) including
any paid or unpaid overtime?

123
462 D. Gallie, H. Russell

Table 6 continued

Question wording Response category

My job requires that I work very hard 1 Agree Strongly

I never seem to have enough time to get everything 2 Agree
done in my job 3 Neither agree nor disagree
4 Disagree
5 Disagree strongly
My wage or salary depends on the amount of effort I put 1 Not at all true,
into my work 2 a little true
3 quite true
4 very true
I can decide the time I start and finish work 1 not at all true
2 a little true
3 quite true
4 very true
My health or safety is at risk because of my work 1 not at all true
2 a little true
3 quite true
4 very true
My job is secure 1 not at all true
2 a little true
3 quite true
4 very true
How often does your work involve 1 Never
… working in the evenings or nights 2 Less than once a month
… having to work overtime at short notice 3 Once a month
4 Several times a month
5 Once a week
6 Several times a week
7 Every Day
How often does your work involve working at 1 Never
weekends? 2 Less than once a month
3 Once a month
4 Several times a month
5 Every week
Around how large a proportion of the household income 1 None
do you provide yourself? 2 Very small
3 Under a half
4 About a half
5 Over a half
6 Very large
7 All
Thinking about the total amount of time people in your Hours
household spend on housework at your home, about
how many hours are spent doing housework
… during a typical weekend
… on a typical weekday
About how much of (total housework time) does your 1 none/almost none
husband/wife/partner spend on housework or up to a quarter
… during a typical weekend 0 more than a quarter up to all
… on a typical weekday

123
Work-Family Conflict and Working Conditions 463

Table 7 Country level models of work-family conflict

GB DK SE NO DE FR NL
B B B B B B B

Constant 1.34** 1.12** 1.16** 1.15** .77** .55(*) .65*

Female .08 .21* .28** .16* .22** .28** .45**
Child under 6 years -.16(*) .35** .09** .18** .09 .11 .04
Child 6–17 years -.07 .23** .04 .15** .14* .12(*) .07
Weekly housework hours .00(*) .00* .00 .00 .00 .01* .00
for HH
Partner employed .07 .06 -.01 -.25* -.14 .26(*) .03
% Financial contribution -.01 .00 .02 .06* -.01 .07* .07*
to HH
Partner \ 1/4 housework -.03 .02 .03 .01 -.16* -.11 .07
weekday
Partner \ 1/4 housework .01 -.10 -.08 -.12(*) .01 .01 -.23**
weekend
Lower professional- -.02 -.09 -.13(*) -.14* -.06 -.13 .03
managerial
Routine non-manual -.35** -.06 -.20* -.25** -.14(*) -.17(*) -.16(*)
Skilled manual -.38* -.34** -.22* -.28** -.22* -.52** -.12
Non-skilled -.48** -.35** -.25* -.19* -.10 -.58** -.23*
Work hours per week .01(*) .00 .02** .00 .01** .00 .01**
Unsocial hours index .18** .20** .12** .19** .10** .22** .08**
Task discretion index .00 .00 -.03* -.01 .05** .00 .01
Can decide time -.04 -.01 -.01 -.01 -.04 -.02 -.04
start/finish work
Job pressure index .25** .23** .20** .17** .28** .33** .23**
Health/safety at risk .18** .24** .14** .13** .13** .09** .12**
Job is secure -.10* -.09** -.04 -.05* -.05* -.01 -.06*
Partner’s work hours .00 .00* -.01(*) .00 .00 .00 .00
Partner unsocial hours -.02 .02 .01 .02 .07* .02 .04
Adjusted R2 (family vars) .036 .082 .041 .069 .050 .037 0.080
Adjusted r2 (full model) .305 .382 .273 .330 .362 .351 0.263
N 512 484 648 667 617 577 541

** \ .005, * \ .05, (*) \ .10

123
464 D. Gallie, H. Russell

Table 8 Country models of work-family conflict by gender

Alla Alla NO NO SE SE DK DK
Men Women Men Women Men Women Men Women

Constant 1.43** .87** 1.32 1.07** 1.62** 1.15** 1.28** 1.20*

Child under 6 years 0.10* .12** .18 .21* .11 .13 .27* .38**
Child 6–17 years 0.06(*) .12** .18 .13 .08 .03 .19* .27**
Weekly housework hours 0.00** .00** .00 .00 .00 .00 .00(*) .00
for HH
Partner employed -0.06 .00 -.41** -.08 .01 .16 .02 .19
HH income -0.02 .06** .05 .07(*) -.02 .05 -.04 .03
contribution (%)
Partner\1/4 housework 0.03 .00 -.02 .03 .03 .02 .17 .00
weekday
Partner\1/4 housework 0.01 -.10** .00 -.15(*) .08 -.15 .04 -.13
weekend
Lower professional- -0.12** -.01 -.18* -.10 -.17(*) -.05 -.05 -.15
managerial
Routine non-manual -0.24** -.11* -.26* -.20(*) -.52** -.03 .02 -.17
Skilled manual -0.30** -.38** -.30** -.34 -.27* .04 -.24(*) -.89*
Non-skilled -0.34** -.17* -.29** .05 -.37** -.16 -.17 -.56*
Work hours per week 0.01** .01** .00 .00 .01(*) .02** -.01 .01
Unsocial hours index 0.17** .16** .21** .17** .20** .05 .24** .16**
Task discretion index 0.01 .01 .00 -.01 -.04* -.02 .02 -.01
Can decide time -0.02 -.03* .02 -.04 .00 -.03 .03 -.03
start/finish work
Job pressure index 0.22** .25** .15** .19** .17** .22** .27** .17**
Health/safety at risk .15** .14** .12* .15** .12** .15** .22** .27**
Jon is secure -.07** -.05** -.08* -.01 -.04 -.05 -.14** -.06
Partner’s work hours 0.00 .00 .00 .00 -.01 -.01(*) .00 .00
Partner unsocial hours 0.02 .02(*) .05(*) -.01 -.05 .06 .00 .03

123
Work-Family Conflict and Working Conditions 465

Table 8 continued
GB GB France France NL NL DE DE
Men Women Men Women Men Women Men Women

Constant .86(*) 2.11** 1.21* .23 .91* .50 .68* 1.15**

Child under 6 years -.04 -.09 .11 .20 .03 .12 .14 .05
Child 6–17 years -.08 .07 -.07 .39** -.05 .20* .12 .17*
Weekly Housework hours .01* .00 .01* .00 .00 .00 .00 .00(*)
for HH
Partner employed -.14 -.04 .30 .14 -.27(*) .41* -.03 -.33(*)
HH income contribution % .02 -.06 -.02 .13* .08(*) .09* -.06 .04
Partner\1/4 housework -.10 .21 -.07 -.03 -.12 .17(*) -.16 -.16
weekday
Partner\1/4 housework .16 -.14 .30 -.15 -.17 -.27** -.09 .04
weekend
Lower professional- .04 -.07 -.21(*) .05 .09 .04 .01 -.24(*)
managerial
Routine non-manual -.35* -.35* -.06 -.06 -.32* .02 -.08 -.27*
Skilled manual -.86* -.34* -.54* -.65(*) -.13 -.50(*) -.16 -.39(*)
Non-skilled -.68** -.45* -.65** -.32(*) -.46* .24 -.07 -.20
Work hours per week .01* .00 .01(*) .00 .01* .02** .01** .01*
Unsocial hours index .22** .16* .16** .27* .08* .09* .08* .12**
Task discretion index .02 -.02 .02 -.01 .01 .01 .07** .05**
Can decide time start/finish -.02 -.04 -.06 .00 -.04 -.08* -.01 -.08*
work
Job pressure index .34** .23** .26** .37** .13* .34** .28** .22**
Health/safety at risk .18* .16* .10* .10* .17** .10(*) .16** .16**
Jon is secure -.12* -.09(*) -.04 .01 -.04 -.09* -.03 -.04
Partner’s work hours .01 .00 -.01 -.01 .00 -.01* .00 .00
Partner unsocial hours -.12* .09 .00 .02 .10* -.02 .03 .13**
a
The pooled country model includes controls for country dummies
** \ .005, * \ .05, (*) \ .10

Table 9 Heckman selection model: women

Coefficient Significance

Work-family conflict
Constant 1.463 .000
Denmark -.047 .506
Norway -.188 .005
Sweden -.110 .107
Germany .041 .555
NL -.054 .445
France .200 .010
Partner employed -.007 .944
Child under 6 years .081 .057

123
466 D. Gallie, H. Russell

Table 9 continued

Coefficient Significance

Child 6–17 years .083 .027

Weekly housework hours for HH .004 .000
Percentage of financial contribution to HH .006 .172
Partner \1/4 housework weekday -.058 .135
Partner \1/4 housework weekend -.087 .023
Lower professional-managerial -.073 .108
Routine non-manual -.209 .000
Skilled manual -.425 .000
Non-skilled -.305 .000
Discretion index .014 .106
Decide start/finish time -.033 .033
Job pressure .280 .000
Health & safety at risk .107 .000
Hours of work .001 .002
Unsocial hours Index .188 .000
Job is Secure -.051 .000
Partner’s work hours -.003 .170
Partner unsocial hours .013 .399
Selection into employment
Age .286 .000
Age squared -.004 .000
Denmark .370 .000
Norway .408 .000
Sweden .451 .000
Germany -.382 .000
NL -.190 .058
France .049 .629
Years of education -.003 .418
No of children under 18 years -.264 .000
Wald test of independent equations (rho = 0): v2(1) = 2.8 Prob [ v2 = 0.0941

References

Askenazy, P., Cartron, D., de Coninck, F., & Gollac, M. E. (2006). Organisation et intensité du travail.
Toulouse: Editions Octares.
Bruning, G., & Plantenga, J. (1999). Parental leave and equal opportunities in eight European countries.
Journal of European Social Policy, 19, 195–209.
Byron, K. (2005). A meta analytic review of work family conflict and its antecedents. Journal of Vocational
Behavior, 67(2), 169–198.
Crompton, R., Lewis, S., & Lyonette, C. (2007). Women, men, work and family in Europe. Palgrave:
Macmillan.
Crompton, R., & Lyonette, C. (2007). Are we all working too hard? Women, men and changing attitudes to
employment. In A. Park, J. Curtice, K. Thomson, M. Philips, & M. Johnson (Eds.), British social
attitudes. Perspectives on a changing society. The 23rd report. London: Sage Publications.

123
Work-Family Conflict and Working Conditions 467

Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Cambridge: Polity Press.
Folbre, N., & Bittman, M. (Eds.). (2004). Family time: The social organization of care. London: Routledge.
Gallie, D. (2007a). Production regimes and the quality of employment in Europe. Annual Review of
Sociology, 33, 5.1–5.20.
Gallie, D. (Ed.). (2007b). Employment regimes and the quality of work. Oxford: Oxford University Press.
Gershuny, J. (2000). Changing times: work and leisure in post industrial society. Oxford: Oxford University
Press.
Gornick, J. C., & Meyers, M. K. (2003). Families that work: Policies for reconciling parenthood and
employment. New York: Russell Sage Foundation.
Gornick, J., Meyers, M., & Ross, K. (1997). Supporting the employment of mothers: Policy variation across
fourteen welfare states. Journal of European Social Policy, 7(1), 45–70.
Green, F. (1999). It’s been a hard day’s night: The concentration and intensification of work in late 20th
century Britain. Department of Economics, University of Kent, Studies in Economics 99/13.
Green, F., & McIntosh, S. (2001). The intensification of work in Europe. Labour Economics, 8, 291–308.
Hochschild, A. (1990). The second shift. London: Piatkus.
Jacobs, J. A., & Gerson, K. (2001). Overworked individuals or overworked families? Explaining trends in
work, leisure and family time. Work and Occupations, 28(1), 41–63.
Jacobs, J., & Gerson, K. (2004). The time divide: work, family, and gender inequality. Cambridge: Harvard
University Press.
Karesek, R., & Theorell, T. (1990). Healthy work. Stress, productivity and the reconstruction of work life.
New York: Basic Books.
Lewis, J. (1992). Gender and the development of welfare regimes. Journal of European Social Policy, 2(3),
159–173.
Lewis, J. (2006). Men, women, work, care and policies, review essay. Journal of European Social Policy,
16(4), 387–392.
Pichler, F. (2008). Work-life balance: Shortcomings in the contemporary measurement of WLB in large-
scale surveys. Social Indicators Research. doi:10.1007/s11205-008-9297-5.
Plantenga, J., & Remery, C. (2005). Reconciliation of work and private life: A comparative review of thirty
European countries. Luxembourg: Office for Official Publications of the European Communities,
European Commission.
Scherer, S., & Steiber, N. (2007). Work and family in conflict? The impact of work demands on family life.
In D. Gallie (Ed.), Employment regimes and the quality of work. Oxford: Oxford University Press.
Strandh, M., & Nordenmark, M. (2006). The Interference of paid work with household demands in different
social policy contexts: Perceived work-household conflict in Sweden, the UK, the Netherlands,
Hungary, and the Czech Republic. The British Journal of Sociology, 57(4), 597–617.
Tahlin, M. (2007). Skills and wages in European labour markets. In D. Gallie (Ed.), Employment regimes
and the quality of work. Oxford: Oxford University Press.
Theorell, T. (1998). Job characteristics in a theoretical and practical health context. In C. L. Cooper (Ed.),
Theories of organizational stress. Oxford: Oxford University Press.
White, M., Hill, S., McGovern, P., Mills, C., & Smeaton, D. (2003). High-performance management
practices, working hours and work-life balance. British Journal of Industrial Relations, 41(2), 175–
195.

123