Journal of Organizational Behavior

J. Organiz. Behav. 20, 879±895 (1999)

The e€ects of job stressors on marital

satisfaction in Finnish dual-earner couples
SAIJA MAUNO* AND ULLA KINNUNEN
University of JyvaÈskylaÈ, Family Research Unit, JyvaÈskylaÈ, Finland

Summary The focus on the present study was to test a mediational model appropriate for explain-
ing the e€ects of psychosocial job stressors, i.e., job insecurity, job autonomy, time
pressures at work, leadership relations and work±family con¯ict, on marital satisfaction
via job exhaustion and psychosomatic health. The study was carried out among 215
married or cohabiting dual-earner couples. The proposed model was tested through
structural equation analysis (LISREL). The results indicated that the job stressors,
except for job autonomy, spilled over into marital satisfaction via job exhaustion and
psychosomatic health for both men and women. However, no empirical support was
found for the crossover of job stressors between partners, signifying that job stressors
experienced by one partner did not in¯uence the marital well-being of the other.
Copyright # 1999 John Wiley & Sons, Ltd.

Introduction

There is empirical evidence for a relationship between experiences at work and marital func-
tioning. Accordingly, positive work experiences (e.g., job satisfaction) are related to positive
marital functioning, whereas negative work experiences (e.g., work stress) are associated with
marital dysfunction (see Barling, 1990). In the psychological literature the mechanism which
combines these two areas of life is commonly de®ned as `spillover', whereby reactions experi-
enced in the work domain are transferred to another, nonwork domain, and vice versa (see
e.g., Lambert, 1990; Leiter and Durup, 1996; Zedeck and Mosier, 1990).
Despite the general agreement on spillover between the domains of work and marriage,
previous studies (e.g., Barling and MacEwen, 1992; Hughes and Galinsky, 1994; Hughes,
Galinsky and Morris, 1992) have provided only limited evidence for the spillover relationship
between speci®c psychosocial job stressors and marital functioning. In other words, the processes
by which psychosocial job stressors a€ect marital functioning have not yet been suciently
speci®ed. The focus of the present study is to gain further insight into these mediational
processes.

* Correspondence to: Saija Mauno, University of JyvaÈskylaÈ, Family Research Unit, P.O. Box 35, FIN-40351 JyvaÈskylaÈ,
Finland. Tel: ‡358-14-60-2891. Fax: ‡358-14-60-2841. E-mail: [email protected].®
Contract grant sponsor: Finnish Work Environment Fund.
Contract grant number: 95196.

CCC 0894±3796/99/060879±17$17.50
Copyright # 1999 John Wiley & Sons, Ltd. Accepted 29 July 1998
880 S. MAUNO AND U. KINNUNEN

Recent studies have shown that negative work experiences, usually conceptualized and
measured as work stress or stressors, exert an in¯uence on marital functioning via their e€ect on
individual well-being (see Barling and MacEwen, 1992; Higginbottom, Barling and Kelloway,
1993; Kelloway and Barling, 1995; MacEwen, Barling and Kelloway, 1992; Matthews, Conger
and Wickrama, 1996). Thus, individual well-being functions as a mediator between work experi-
ences and marital functioning. In these studies individual well-being has been used to refer to
various factors covering both occupational (e.g., job satisfaction, job exhaustion) and overall
well-being (e.g., psychosomatic health, depressive symptomatology), while marital functioning
has been used to refer, for example, to the quality of interaction (warmth and supportiveness),
psychological aggression, divorce propensity, and marital satisfaction.
The purpose of the present study is to test a mediational model appropriate for explaining the
e€ects of job stressors on marital satisfaction among dual-earner couples in Finland. Our
proposed model, developed on the basis of previous research and based on the spillover
hypothesis, predicts that job stressors will be related directly to well-being in the work domain
and indirectly to overall and marital well-being. In line with Warr's (1987) notion, we
expected that job stressors would initially a€ect context-speci®c well-being in the work domain
and then context-free overall well-being and ®nally family well-being. Hence, the e€ects of
job stressors on marital well-being are assumed to be mediated through occupational and
overall well-being.
The participants of the study were married or cohabitating couples, which enabled us to study
job stress crossover between the partners. Crossover describes how stress and strain experienced
by a partner a€ects the behavior and well-being of the other (Bolger et al., 1989; Jones and
Fletcher, 1993, 1996; Pittman, Solheim and Blanchard, 1996; Westman and Etzion, 1995). We
also expected the crossover processes to occur indirectly, so that, ®rst, job stressors reduce one's
occupational well-being and second, one's overall well-being, which in turn impairs the marital
well-being of one's spouse.
We concentrated on psychosocial job stressors, which are related to the content and processes of
the employee's job (Caplan, 1985). The following job stressors, which are also identi®ed in widely
known stress models, were examined: job insecurity, job autonomy, time pressures at work,
leadership relations, and work±family con¯ict (see e.g., Cooper and Marshall, 1976; Hurrell and
Murphy, 1992; Murphy, 1995). Well-being at work was described by emotional exhaustion which
provides a general measure of the amount of stress and tedium a person is experiencing (Maslach
and Jackson, 1986). On the basis of stress theory, overall well-being was described by psycho-
somatic symptoms.
Job insecurity refers to the amount of uncertainty a person has about his or her job continuity
(see Caplan et al., 1980; Greenhalgh, 1982; 1991, Hartley et al., 1991). Jacobson (1991a, 1991b)
argues that job insecurity is a stressful experience, because it concerns the future; an employee
does not know whether (s)he will actually lose his/her job, and this uncertainty restricts coping
processes available in a stressful situation. As a result, job insecurity may also reduce the well-
being of an employee; this has been supported by recent empirical studies. Fear of losing one's
job has been related to burnout among teachers (McGrath, Houghton and Reid, 1989) and
transport corporation employees (Dekker and Schaufeli, 1995). In addition, job insecurity has
been associated with psychological distress (Catalano, Rook and Dooley, 1986; Dekker and
Schaufeli, 1995) as well as with psychosomatic (Kinnunen and NaÈtti, 1994) and ill-health
symptoms (Kuhnert, Sims and Lahey, 1989).
In this study, job autonomy and time pressures at work represent Karasek's (1979; 1989;
Karasek and Theorell, 1990) widely studied idea of job strain, according to which highly stressful
jobs are characterized by high demands and low control. Job autonomy or control over one's

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
 EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION 881

work refers to task authority, which involves the employee's freedom to determine how work gets
done, setting one's own work goals and using one's skills at work, and having a contribution to
decision making. Many studies have indicated that low job autonomy is associated with psycho-
logical distress (Daniels and Guppy, 1994; Schaefer and Moos, 1993) and reduced well-being, for
example, burnout (McKnight and Glass, 1995; Turnipseed, 1994). Working conditions that
provide a low degree of control might also constitute an increased health risk for employees
(Gaillard and Wientjes, 1994; Karasek and Theorell, 1990).
Time pressures at work can be de®ned as having too much to do in a limited amount of time.
Severe pressures of time have been identi®ed as a source of work stress, for example, for preschool
teachers (Kelly and Barthelsen, 1995), physicians (Richardsen and Burke, 1991; 1993) and health
care personnel (LindstroÈm, 1992). Studies have shown that high work pressures (e.g., time
shortages), work load, and schedule considerations are signi®cantly linked to emotional exhaus-
tion and burnout (Dekard, Meterko and Field, 1994; Leiter and Durup, 1996; Turnipseed, 1994)
as well as to an increase in psychosomatic complaints (Houtman et al., 1994).
The quality of leadership relations encompasses, for example, an employee's relationship with
his/her supervisors, and particularly both the quality and the quantity of supervisory support. A
low level of supervisor support has been related to psychosocial stress (Jones-Johnson and
Johnson, 1992) and to burnout and emotional exhaustion (Himle, Jayaratne and Thyness, 1989;
Turnipseed, 1994). Working conditions supplying little social support may also contribute
negatively to an employee's health (Gaillard and Wientjes, 1994).
Work±family con¯ict re¯ects the interference of work with family activities, for example,
where work prevents the performance of duties at home and spending time with one's family (see
Frone, Russell and Cooper, 1992a; Gutek, Searle and Klepa, 1991). Con¯ict between work and
family may be perceived as a form of role con¯ict; such a con¯ict could be regarded as a source of
job stress, leading to reduced psychological (Frone, Russell and Barnes, 1996; Frone et al., 1992a;
Stewart and Barling, 1996; Thomas and Ganster, 1995) and physical (Frone et al., 1996; Thomas
and Ganster, 1995) well-being. In fact, work±family con¯ict has predicted job exhaustion
and burnout (Bacharach, Bamberger and Conley, 1991; Burke, 1993), which in turn has been
associated with reduced marital satisfaction (Leiter and Durup, 1996).
Some studies have focused on the indirect relationship between job stressors and marital well-
being. Barling and MacEwen (1992) found that depression mediates the relationship between job
insecurity and marital satisfaction. Matthews et al. (1996) in turn indicated that work±family
con¯ict a€ected marital quality and stability via greater psychological distress, marital hostility
and lowered marital warmth and supportiveness. However, Widrich and Ortlepp (1994) found no
empirical support for the mediating processes (i.e., job satisfaction) between job stress and
marital functioning.
The possibility of a direct relationship between job stressors and marital well-being has been
investigated more often. Job insecurity has been associated directly with increased marital
tension (Hughes and Galinsky, 1994), and with decreased marital satisfaction (Larson, Wilson
and Beley, 1994). Stets (1995) indicated that reduced job autonomy in¯uenced the marital
relationship: individuals responded to reduced autonomy at work by controlling their spouses at
home or by becoming depressed and consequently withdrawing at home. Several studies have
shown that work±family con¯ict is also associated directly with reduced marital well-being (see
e.g., Campbell and Snow, 1992; Coverman, 1989; MacEwen and Barling, 1988).
The framework of the present study was a mediational model linking job stressors to
marital satisfaction via individual well-being (Figure 1). First, we hypothesized that job stressors
(i.e., job insecurity, low job autonomy and high time pressures at work, poor leadership
relations and work±family con¯ict) result in impaired context-speci®c occupational well-being

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
882 S. MAUNO AND U. KINNUNEN

Figure 1. The proposed theoretical model linking job stressors to marital satisfaction via individual
well-being

(i.e., job exhaustion), which in turn is associated with reduced context-free overall well-being
(i.e., psychosomatic symptoms). Reduced well-being at these two levels was expected to spill
over into family well-being (i.e., marital satisfaction). We assumed that a similar model would
be adequate for both men and women, i.e., no gender di€erences in this mediational process
was assumed, since there is no evidence to show that the spillover mechanisms di€er between the
sexes.
Second, we assumed that job stressors experienced by one partner would be associated not
only with his/her reduced well-being, but also with the reduced well-being of his/her spouse
(see Figure 1). We hypothesized that job stressors lead to reduced psychosomatic well-being via
job exhaustion and that reduced psychosomatic well-being for one partner is in turn re¯ected in
the marital well-being of the other. We expected to ®nd this crossover only in the domain of
marital satisfaction since the marital relationship constitutes a sphere which is commonly shared
by both partners.
This proposed crossover model was also assumed to concern men and women equally, i.e., the
crossover process would proceed from husband to wife as well as from wife to husband. This
hypothesis is in line with the few studies (Matthews et al., 1996; Westman and Etzion, 1995;
Zedeck et al., 1989) which have found the nature of the crossover process to be bidirectional.
However, there are also studies which contradict our hypothesis by suggesting that the crossover
process is primarily unidirectional, that is, from husband to wife (see Jones and Fletcher, 1993;
Pittman et al., 1996; Rook, Dooley and Catalano, 1991). Our hypothesis was based on the fact
that in Finland the roles of women and men are relatively equal compared to the situations in
many other countries; for example, women constitute more than half of the total working
population and most women are employed full-time (Kauppinen-Toropainen, 1993). This means

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
 EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION 883

that work may have rather similar e€ects in both men's and women's lives (see also Barnett, 1997;
Barnett and Brennan, 1997; Windle and Dumenci, 1997).

Method
Participants
The participants in the study were 215 married (88 per cent) or cohabitating (12 per cent) dual-
earner partners in Finland. Data for the study were obtained as part of a larger research project
`Job Insecurity and Well-Being', in which the participants were employed in three di€erent
economic domains in Central Finland: the public sector (municipal social and health care,
N ˆ 2054), manufacturing for export (a factory, N ˆ 472) and supplying goods and services to
the domestic markets (a bank, N ˆ 319, and a supermarket, N ˆ 117). All those employed at the
supermarket participated in the project, whereas the other organizations provided a list of
employees from which random samples were selected. Of the 979 questionnaires distributed, 518
were returned, resulting in a response rate of 53 per cent. About 72 per cent (215/299) of the
couples in the larger research project participated in this study. One partner from each couple
worked in one of the three above-mentioned economic domains. The total distribution among
the domains was as follows: 48 per cent in the public sector, 30 per cent in the domestic market,
and 22 per cent in manufacturing for export. The demographic characteristics of these dual-
earner partners are presented in Table 1.
The men were older than the women: the average age of the men was 43.9 and of the women
42.6 years (t 214† ˆ 5.66, p 5 0.001). Of the couples 35 per cent had no children under 18 years
living at home, whereas under 30 per cent had one child, 25 per cent two, and 13 per cent three or
more children. Thus, the modal number of children living at home was 1±2 in these families. The
age of the youngest child living at home varied as follows: 30 per cent were less than seven years of
age, 26 per cent between 7 and 12 years, 25 per cent between 13 and 18 years, and 19 per cent were
older than 18 years of age. More women had completed a vocational college education, whereas
the men had usually a more manually oriented vocational school education. Thus, the women
were more often employed as lower non-manual workers, and the men as manual workers. In
addition, the men were employed in upper non-manual occupations and had a position of
leadership more often than the women. Both the men and the women in the sample were usually
employed full-time and had a regular day shift. The greater part of the participants worked on a
permanent basis and had a long work history with their present employer.

Procedure

The data were collected in February 1996 by means of two structured questionnaires. The ®rst, an
18-page questionnaire, was distributed to those employees working for the four employers
described above. The second, a seven-page questionnaire, was directed to the employees' spouses
in the case he/she was employed and it was delivered by the employee. The questionnaires were
code-numbered to match the partners correctly. Despite this code numbering the participants
remained unidenti®ed as both questionnaires were answered anonymously. Each partner was
advised to ®ll in the questionnaire independently. Respondents returned their questionnaires in
closed envelopes to the mailbox on the work premises. The questionnaires consisted of several

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
884 S. MAUNO AND U. KINNUNEN

Table 1. The demographic characteristics for dual-earner partners (n ˆ 215 couples)

Characteristic (%) Men Women w2
Age (years)
18±24 1 1 3.59
25±34 10 13
35±44 39 42
45±54 44 40
55±64 6 4
Vocational education
None 25 15 11.02*
Vocational school 31 32
Vocational college 28 41
Higher 16 12
Socioeconomic status
Manual workers 41 12 77.09{
Lower non-manual workers 28 70
Upper non-manual workersa 31 17
Leadership position
Not superior 60 83 26.68{
Superior 40 17
Working hours per week
Part time job (535 h/week) 8 11 1.31
Full time job (435 h/week) 92 89
Working schedule
Regular day shift 63 71 2.69
Other shifts 37 29
Working on a permanent basis
Not permanent 3 5 1.04
Permanent 97 95
Job tenure (years)
0±4 9 15 5.43
5±10 24 26
11±20 37 38
Over 20 30 22
* p 5 0.05, { p 5 0.001.
Note. a includes entrepreneurs; for men 8 per cent and for women 1 per cent.

questions; however only those which are relevant to the present report are discussed below in
detail.

Measures

The study variables were distributed into three groups as follows: the predictor variables (i.e., job
stressors), the mediator variables, which assessed individual well-being (i.e., job exhaustion,
psychosomatic symptoms), and the outcome variable (i.e., marital satisfaction). Composite
variables were created by averaging the respective items and scoring the scales so that a high score
represented a high level of the measured construct.

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
 EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION 885

Job stressors
The ®rst predictor variable, job insecurity was measured using a global de®nition, which focuses
on the threat of job loss or job continuity (see Greenhalgh, 1982; Hartley et al., 1991). The
measure included four items which assessed concern (e.g., `I am worried about the possibility of
being ®red') and certainty (e.g., `I am certain that my job will continue for a long time') about job
retention. The couples responded on a 5-point agree/disagree scale. The Cronbach alpha
coecients for the scale were 0.74 for the men and 0.76 for the women.
The measurement of job autonomy and time pressures at work were based on Karasek's (1979;
Karasek and Theorell, 1990) job strain model. Job autonomy was measured by seven items which
concerned the employee's in¯uence at work, including the following factors: division of labor,
working pace and working methods. The couples answered on a 4-point scale. The alphas for the
scale were 0.87 for the men and 0.83 for the women. Time pressures at work were measured by ®ve
items (e.g., `I have to rush at work to get everything done'). The couples responded on a 4-point
scale (1 ˆ not at all, 4 ˆ much). The alphas for the scale were 0.80 and 0.82 for the men and
women, respectively.
Leadership relations were assessed using four items which stressed feedback and support from
supervisors (e.g., `I can get help and support from my supervisor when I experience diculties at
work'). The respondents answered on a 5-point agree/disagree scale. The alphas for the scale were
0.84 for both sexes. The items were based on the study `The Quality of Working Life in Finland'
(see Kinnunen and NaÈtti, 1994).
The measurement of work±family con¯ict was derived from Frone et al. (1992a; 1992b) and
Brett and Yogev (1985). The measure consisted of three items, which assessed the degree to which
a respondent's work interferes with his/her family life (e.g., `How often does your job or career
interfere with your responsibilities at home, such as yard work, cooking, cleaning, repairs, or
child care?'). The respondents answered on a 5-point frequency-based response scale (1 ˆ never,
5 ˆ always). The alphas for the scale were 0.81 for the men and 0.77 for the women.

Individual well-being
The ®rst mediator variable, job exhaustion, was measured in a way congruent with Maslach and
Jackson's (1981; 1986) burnout model. However, only six items concerning feelings of fatigue
that develop as one's emotional energies become drained at work were used (e.g., `I feel like I'm at
the end of my tether'). The couples responded with ®ve response options (1 ˆ never,
5 ˆ always). The alphas for job exhaustion were 0.85 for the men and 0.90 for the women.
The second mediator variable, psychosomatic symptoms, was used as a measure of overall well-
being. The symptoms (e.g., tension, strain, palpitations or irregular heartbeat, headache) were
rated on a 10-item scale, with ®ve response options (1 ˆ never, 5 ˆ all the time). The measure
was derived from the study `The Quality of Working Life in Finland' (see Kinnunen and NaÈtti,
1994). The alpha coecients for the scale were 0.88 for the men and 0.90 for the women.

Marital satisfaction
The outcome variable, marital satisfaction, was evaluated using Fowers and Olson's (1993)
Enrich Marital Satisfaction Scale (ENRICH) in a shortened form. The scale comprised seven
items (e.g., `I am very happy with how we manage our leisure activities and the time we spend
together') which were answered on a 5-point agree/disagree scale. The alphas for the scale were
0.78 for both sexes.

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
886 S. MAUNO AND U. KINNUNEN

Data analysis

The matched responses of each participant and those of his/her partner were analysed using the
LISREL model, which permits the speci®cation and estimation of the model hypothesized to
account for the data. The proposed model was estimated using a covariance matrix (LISREL
8.12; JoÈreskog and SoÈrbom, 1996) with the method of estimation being that of maximum
likelihood (the variables were normally distributed). The chi-square statistic, the goodness-of-®t
index (GFI) and root mean square error of approximation (RMSEA), comparative ®t index
(CFI), and normed ®t index (NFI) were used to test the adequacy of the model.
For two speci®c reasons the models were estimated simultaneously for both sexes. First, it was
not possible to investigate the crossover process between the partners using separate models for
males and females. Second, the married or cohabiting couples had to be considered as a
dependent sample, and hence analysed simultaneously. For these speci®c reasons, the theoretical
fully-mediated model, presented in Figure 1, was analysed simultaneously for both sexes.
However, even though a model ®ts the data it does not necessarily mean it is a correct model.
Two other types of models were estimated by LISREL: the partially mediated model and the non-
mediated model. Therefore, the fully mediated model was tested ®rst, followed by the partially
mediated model tested against the fully mediated model. In the partially mediated model, one
parameter indicating a direct link between job exhaustion and marital satisfaction was added
to the analysis. Second, the non-mediated model was tested against the partially mediated model.
In the non-mediated model, the parameter between psychosomatic symptoms and marital satis-
faction was deleted, and only the parameters between job exhaustion and psychosomatic
symptoms, and between job exhaustion and marital satisfaction, were estimated.
In addition, the parameters for the sexes in each of the models described above were ®rst
estimated as unequal, and then constrained to be equal. Following this, the models were
compared using chi-square statistics in order to ®nd out which one of the models best ®tted the
data. If adding a parameter resulted in a signi®cant decrease in chi-square value, it was taken to
indicate that the parameter signi®cantly improved the ®t of the model and was therefore included.
In each tested model the error covariances between a husband's job stressors, and respectively the
error covariances between a wife's job stressors, were set free.

Results
Descriptive results
The means, standard deviations and correlations of all the study variables are presented
separately for the men and the women in Table 2. In addition, the intercorrelations of the study
variables between the partners are presented in Table 3. Gender di€erences emerged in three of
the variables studied. First, the men perceived more autonomy in their jobs than did the women
(t 211† ˆ ÿ4.59, p 5 0.001). Second, the men had somewhat more con®dence in the continuity
of their jobs than the women (t 209† ˆ 2.6, p 5 0.01). Third, the women su€ered more from
psychosomatic symptoms than did the men (t 210† ˆ 2.64, p 5 0.01).
As seen in Table 2, the intercorrelations were in the expected direction, with job stressors
correlating with individual well-being (i.e., job exhaustion and psychosomatic symptoms), which
in turn was associated with marital satisfaction among both sexes. Furthermore, the job stressors

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
Copyright # 1999 John Wiley & Sons, Ltd.

Table 2. Intercorrelations among the study variables for men and women

EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION

Variable Men Women 1 2 3 4 5 6 7 8
M S.D. M S.D.
1. Job insecurity 2.33 0.91 2.46 0.94 ± ÿ0.27 ÿ0.01 ÿ0.29 0.00 0.13 0.25 ÿ0.08
2. Job autonomy 2.55 0.69 2.27 0.60 ÿ0.24 ± 0.15 0.21 0.12 0.12 ÿ0.01 0.06
3. Time pressures at work 2.36 0.65 2.37 0.65 0.05 ÿ0.17 ± ÿ0.08 0.43 0.55 0.36 ÿ0.15
4. Good leadership relations 3.29 0.91 3.26 1.04 ÿ0.04 0.43 ÿ0.18 ± ÿ0.12 ÿ0.20 ÿ0.19 0.05
5. Work±family con¯ict 2.76 0.87 2.68 0.81 ÿ0.02 ÿ0.13 0.37 ÿ0.19 ± 0.57 0.42 ÿ0.21
6. Job exhaustion 2.12 0.73 2.21 0.83 0.09 ÿ0.24 0.57 ÿ0.33 0.53 ± 0.73 ÿ0.16
7. Psychosomatic symptoms 1.75 0.58 1.91 0.67 0.16 ÿ0.24 0.39 ÿ0.31 0.38 0.72 ± ÿ0.24
8. Marital satisfaction 3.84 0.68 3.74 0.76 0.00 0.15 ÿ0.03 0.17 ÿ0.18 ÿ0.19 ÿ0.24 ±
J. Organiz. Behav. 20, 879±895 (1999)

Note. Coecients above the diagonal are for men, those below the diagonal are for women. For men and women r 5 j 0.19 j , p 5 0.01, r 5 j 0.24 j , p 5 0.001.

887
888 S. MAUNO AND U. KINNUNEN

Table 3. Intercorrelations among the study variables between the partners

Variable For men
1 2 3 4 5 6 7 8
For women
1. Job insecurity 0.12 0.01 0.02 0.01 0.08 0.03 0.10 0.01
2. Job autonomy ÿ0.19 0.08 0.06 0.01 ÿ0.03 ÿ0.00 ÿ0.05 ÿ0.02
3. Time pressures at work 0.13 ÿ0.11 0.09 ÿ0.14 0.13 0.07 0.07 0.07
4. Good leadership relations ÿ0.15 0.07 0.00 0.05 0.02 ÿ0.03 0.00 0.02
5. Work±family con¯ict 0.10 ÿ0.08 0.02 ÿ0.17 0.11 0.09 0.08 ÿ0.06
6. Job exhaustion 0.16 ÿ0.10 0.04 ÿ0.12 0.04 0.08 0.09 ÿ0.01
7. Psychosomatic symptoms 0.17 ÿ0.09 0.07 ÿ0.15 0.05 0.06 0.12 ÿ0.06
8. Marital satisfaction ÿ0.00 0.04 ÿ0.04 0.02 ÿ0.02 ÿ0.01 ÿ0.11 0.52
Note. For men and women: r 5 j 0.15 j , p 5 0.05, r 5 j 0.19 j , p 5 0.01, r 5 j 0.52 j , p 5 0.001.

were associated with one another. These intercorrelations were rather similar in the two sexes,
with one exception. Work±family con¯ict and time pressures at work appeared to be most
strongly linked r ˆ 0.43 and 0.37 for men and women respectively). Also, a high level of job
autonomy was signi®cantly associated with good relations with the leadership (r ˆ 0.21 and 0.43
for men and women respectively) and with a low level of job insecurity (r ˆ ÿ0.27 and ÿ0.24 for
men and women respectively). However, high job insecurity was linked to poor relations with the
leadership for the men (r ˆ ÿ0.29), but not for the women (r ˆ ÿ0.04). Of the associations of
the variables between the couples (see Table 3), only marital satisfaction showed a high
correlation (r ˆ 0.52), with the other variables remaining relatively poorly correlated.

Tests of the proposed model and alternative models

The comparisons of the chi-square statistics for the di€erent models tested are presented in
Table 4. Firstly, these comparisons indicate that each of the models tested could be estimated by
constraining the parameters equal for the sexes. Second, the fully mediated model (i.e., the
proposed model in Figure 1) with equal parameters for the sexes, ®tted the data best
w2 91† ˆ 106.50, p ˆ 0.130, GFI ˆ 0.940, RMSEA ˆ 0.029, CFI ˆ 0.980, NFI ˆ 0.890).
On testing the other potential mediational models, one parameter indicating a direct link
between job exhaustion and marital satisfaction, was added to the analysis in the partially
mediated model. However, adding this parameter did not result in a signi®cantly improved ®t
(w2 1† ˆ 0.60, p ˆ 0.439) compared with the fully mediated model. In the non-mediated model,
the parameter between psychosomatic symptoms and marital satisfaction was deleted, and only
the parameters between job exhaustion and psychosomatic symptoms, and between job
exhaustion and marital satisfaction, were estimated. As a result of this, the model ®t to the data
decreased w2 1† ˆ 9.51, p ˆ 0.002) in this non-mediated model, compared with the partially
mediated model. Taken together, the proposed fully mediated model showed the best ®t to the
data and thus increased con®dence in the validity of the hypothesized model.
The fully mediated theoretical model, however, had to be modi®ed, since there were three
statistically non-signi®cant paths (t-values less than 2.0), which required deleting. Hence, the
crossover paths, i.e., the paths between the husband's psychosomatic symptoms and the wife's
marital satisfaction (t-value ˆ ÿ0.98) had to be deleted, as was the case with the path between
the wife's psychosomatic symptoms and the husband's marital satisfaction. In addition, the path
between job autonomy and job exhaustion turned out to be non-signi®cant (t-value ˆ 0.24). On

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
Copyright # 1999 John Wiley & Sons, Ltd.

Table 4. Goodness-of-®t statistics (chi square) for the tested models

EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION

Unequality/equality of the Fully mediated Partially mediated FM vs. PM Non-mediated PM vs. NM
parameters for the sexes model (FM)a model (PM) model (NM)
Unequal parameters w2 83† ˆ 99. 28, w2 81† ˆ 98.59, w2 2† ˆ 0.59, w2 83† ˆ 108.45, w2 2† ˆ 9.86,
p ˆ 0.110 p ˆ 0.089 p ˆ 0.745 p ˆ 0.032 p ˆ 0.007
Equal parameters w2 91† ˆ 106.50, w2 90† ˆ 105.90, w2 1† ˆ 0.60, w2 91† ˆ 115.41, w2 1† ˆ 9.51,
p ˆ 0.130b p ˆ 0.120 p ˆ 0.439 p ˆ 0.043 p ˆ 0.002
Unequal vs. equal model w2 8† ˆ 7.22, w2 9† ˆ 7.31, w2 8† ˆ 6.96,
p ˆ 0.513 p ˆ 0.605 p ˆ 0.541
Note: The following abbreviations: fully mediated model ˆ FM; partially mediated model ˆ PM; and non-mediated model ˆ NM.
J. Organiz. Behav. 20, 879±895 (1999)

a
corresponding the theoretical model presented in Figure 1.
b
the model best ®tting the data, from which the paths with t-values less than 2.0 were deleted, resulting in the ®nal model presented in Figure 2.

889
890 S. MAUNO AND U. KINNUNEN

Figure 2. The in¯uence of job stressors on marital satisfaction via individual well-being. In order to
simplify the presentation, the correlations among job stressors are not shown.

the contrary, one path, i.e., the path between job insecurity and psychosomatic symptoms, had to
be added according to the modi®cation index (MI for the men 10.64, and for the women 4.94).
Adding this parameter signi®cantly improved the ®t of the model to the data (w2 1† ˆ 15.47,
p ˆ 0.000). Together, these modi®cations resulted in the ®nal model presented in Figure 2
(the standardized solution). The ®t indices showed that the model provided an excellent ®t to the
data (w2 92† ˆ 92.10, p ˆ 0.480, GFI ˆ 0.950, RMSEA ˆ 0.002, CFI ˆ 1.00, NFI ˆ 0.910).
Consistent with our hypothesis, the same model was valid for both men and women, i.e., the
estimated parameters were equal over the sexes in statistical terms.
Of the ®ve job stressors studied, four were directly associated with job exhaustion. Time
pressures at work b ˆ 0.39 and 0.38 for men and women respectively), and work±family con¯ict
(b ˆ 0.40 and 0.35 for men and women respectively) most strongly predicted job exhaustion. In
addition, job insecurity and poor relations with the leadership were linked to job exhaustion. The
workers who experienced high levels of job insecurity, time pressures at work, work±family
con¯ict, and reported having poor relations with the leadership were also likely to experience a
high level of job exhaustion. As proposed, job exhaustion in turn predicted psychosomatic
symptoms (b ˆ 0.73 and 0.74 for men and women respectively). The model accounted for
45±51 per cent of the variance in job exhaustion and 57±58 per cent in the psychosomatic
symptoms. Consistent with the theoretical model, psychosomatic symptoms predicted marital
satisfaction (b ˆ ÿ0.24 and ÿ0.22 for men and women respectively). However, the explained
variance in marital satisfaction was only 5±6 per cent. As seen in Figure 2, there was a positive
reciprocal association between the partners' evaluations of marital satisfaction.
The results were inconsistent with the proposed model for three reasons. First, job autonomy
was not signi®cantly related to occupational well-being. Second, job insecurity was related to

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
 EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION 891

marital satisfaction via two mechanisms, i.e., job exhaustion (b ˆ 0.08) and psychosomatic
health (b ˆ 0.13), meaning that the e€ect of job insecurity on marital satisfaction was mediated
via two separate paths. Third, the model provided no empirical support for the crossover process
between the partners. Individual well-being experienced by one partner did not a€ect the other
partner's marital satisfaction. This ®nding implies that job stressors experienced by one partner,
regardless of gender, were not transmitted to the other as marital satisfaction.

Discussion
The proposed mediational model of job stressors on marital satisfaction received support from
the Finnish dual-earner data. Of the job stressors studied, job insecurity, high time pressures at
work, poor leadership relations, and work±family con¯ict a€ected marital satisfaction through
job exhaustion and the psychosomatic symptoms they created. The results were exactly the same
for both sexes, which means that the process by which mens' and womens' stressors at work a€ect
marital satisfaction is the same. However, the proportion of the variance explained in marital
satisfaction was only 5±6 per cent.
Altogether, the results indicated that the e€ects of job stressors on marital functioning
are primarily indirect. Thus, our results are in line with those of previous studies which have
in particular focused on the indirect nature of the relationship between negative work experi-
ences (e.g., work stress) and marital functioning (see Higginbottom et al., 1993; Kelloway and
Barling, 1995; MacEwen et al., 1992; Matthews et al., 1996). The ®ndings also highlighted the
phenomenon widely referred to as `spillover', whereby reactions and a€ects evoked by work
stressors are transferred to another, nonwork domain (see Barling, 1990; Lambert, 1990; Leiter
and Durup, 1996; Warr, 1987; Zedeck and Mosier, 1990).
Inconsistent with our proposed model and earlier studies (see Jones and Fletcher, 1993;
Karambayaa and Reilly, 1992; Pittman et al., 1996; Rook et al., 1991; Westman and Etzion, 1995;
Zedeck et al., 1989), we found no empirical evidence for the crossover of job stressors between
partners, meaning that the stressors experienced by one partner did not in¯uenceÐdirectly or
indirectlyÐthe marital well-being of the other partner. In line with this result, Jones and Fletcher
(1996) also observed no signi®cant relationship between work stressors experienced by one
partner and the well-being of the other. One interpretation for the non-existence of the crossover
process might be linked to the partners' coping strategies: partners may manage job stressors
e€ectively, for example by discussing job-related problems with each other, which may inhibit
crossover. Partners may also provide each other with emotional support for the relief of negative
emotions and reactions evoked by a stressful work-day (see e.g., Repetti, 1989).
An interesting ®nding was that both the men and the women had a rather similar perception of
their work characteristics in contradiction to several studies which have argued that men's and
women's work experiences are distinguishable (see e.g., Beena and Poduval, 1992; Lai, 1995;
Roxburgh, 1996; Wiersma, 1990). This ®nding is surprising in our data in which men and women
worked not only in di€erent areas of the economy but also in heterogeneous occupational
positions, i.e., occupations were characterized by sex segregation. However, Kauppinen-Toro-
painen, Kandolin and Haavio-Mannila (1988) found that neither the social atmosphere in the
workplace nor job satisfaction varied according to occupational segregation (see also Lambert,
1991; Weeks and Nantel, 1995). These results suggest that gender di€erences in work experiences
are somewhat minor or even diminishing (see Barnett, 1997; Barnett and Brennan, 1997; Windle
and Dumenci, 1997).

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
892 S. MAUNO AND U. KINNUNEN

Contrary to our expectations and to the results of some previous studies (see McKnight and
Glass, 1995; Turnipseed, 1994), job autonomy was related to neither job exhaustion nor psycho-
somatic health. However, job autonomy may have stronger connections with other well-being
indicators, for example, job satisfaction (Greenberger and Grunberg, 1995). The discrepancy
between our and earlier results might also be due to di€erent ways of conceptualizing and
measuring job autonomy.
The cross-sectional study design of the present study seriously limits our knowledge regarding
both the spillover and crossover processes. Although our model was conceptualized in terms of
causes and e€ects, it is recognized that the use of cross-sectional data does not allow us to draw
causal inferences concerning the relationships. We would need longitudinal study designs with
time lags of di€erent lengths between causes and e€ects. One data collection strategy appropriate
for this purpose is the use of daily diary questionnaires (see Bolger et al., 1989; Jones and
Fletcher, 1996; Pittman et al., 1996). Such a strategy may help us in answering the questions
about how prolonged stress or acute job stressors (e.g., daily hassles and disagreements in the
workplace) spill over and are transmitted from one spouse to another.
The shared nature of the cohabitational relationship was also con®rmed in our study by the
high correlations between partners' evaluations of their marital satisfaction. This points to the
possibility that although job-related stressors do not cross over between partners, domestic
stressors, for example, child-care responsibilities, economic stress, and family discord,
experienced by one partner might in¯uence the other partner's a€ective state (see, Jones and
Fletcher, 1996). Consequently, the crossover process occurs predominantly in the family sphere,
which is natural, since the family is a common site for the family members. Instead, the work
sphere is usually a more or less separate and heterogenous area for them. Future research should
also investigate the crossover of family stressors to the work domain, while family stressors
experienced by one partner might be transmitted to the job-related well-being and behavior of the
other partner causing, for example, anxiety at work, withdrawal from interaction in the
workplace, or temporary job dissatisfaction.

References

Bacharach, S. B., Bamberger, P. and Conley, S. (1991). `Work±home con¯ict among nurses and engineers:
mediating the impact of role stress on burnout and satisfaction at work', Journal of Organizational
Behavior, 12, 39±53.
Barling, J. (1990). Employment, Stress and Family Functioning, John Wiley & Sons, Chichester.
Barling, J. and MacEwen, K. (1992). `Linking work experiences to facets of marital functioning', Journal of
Organizational Behavior, 13, 573±583.
Barnett, R. C. (1997). `How paradigms shape the stories we tell: paradigm shift in gender and health',
Journal of Social Issues, 53, 351±368.
Barnett, R. C. and Brennan, R. T. (1997). `Change in job conditions, change in psychological distress,
and gender: a longitudinal study of dual-earner couples', Journal of Organizational Behavior, 18,
253±274.
Beena, C. and Poduval, P. R. (1992). `Gender di€erences in work stress of executives', Psychological
Studies, 37, 109±113.
Bolger, N., DeLongis, A., Kessler, R. C. and Wethington, E. (1989). `The contagion of stress across
multiple roles', Journal of Marriage and Family, 51, 175±184.

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
 EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION 893

Brett, J. M. and Yogev, S. (1985). `Restructuring work for family: how dual-earner couples with children
manage. Special issue: Work and family: Theory, research and applications', Journal of Social Behavior
and Personality, 3, 159±174.
Burke, R. J. (1993). `Toward an understanding of psychological burnout among police ocers', Journal of
Social Behavior and Personality, 8, 425±438.
Campbell, J. L. and Snow, B. M. (1992). `Gender role con¯ict and family environment as predictors of
men's marital satisfaction', Journal of Family Psychology, 6, 84±87.
Caplan, R. D. (1985). `Psychosocial stress at work', Management and Labour Studies, 10, 63±76.
Caplan, R. D., Cobb, S., French, J. R. P. Jr., van Harrison, R. V. and Pinneau, S. R. (1980). Job Demands
and Worker Health: Main E€ects and Occupational Di€erences, Survey Research Center, Institute of
Social Research, University of Michigan, Ann Arbor, Michigan.
Catalano, R., Rook, K. and Dooley, D. (1986). `Labor markets and help-seeking: a test of the employment
security hypothesis', Journal of Health and Social Behavior, 27, 277±287.
Cooper, C. L. and Marshall, J. (1976). `Occupational sources of stress, a review of the literature relating to
coronary heart disease and mental ill health', Journal of Occupational Psychology, 49, 11±28.
Coverman, S. (1989). `Role overload, role con¯ict, and stress: addressing consequences of multiple role
demands', Social Forces, 67, 965±982.
Daniels, K. and Guppy, A. (1994). `Occupational stress, social support, job control, and psychological well-
being', Human Relations, 47, 1523±1544.
Dekard, G., Meterko, M. and Field, D. (1994). `Physicians burnout: an examination of personal,
professional, and organizational relationship', Medical Care, 32, 745±754.
Dekker, S. W. A. and Schaufeli, W. B. (1995). `The e€ects of job insecurity on psychological health and
withdrawal: a longitudinal study', Australian Psychologist, 30, 57±63.
Fowers, B. J. and Olson, D. H. (1993). `ENRICH Marital Satisfaction Scale: a brief research and clinical
tool', Journal of Family Psychology, 7, 176±185.
Frone, M. R., Russell, M. and Barnes, G. M. (1996). `Work±family con¯ict, gender, and health-related
outcomes: a study of employed parents in two community samples', Journal of Occupational Health
Psychology, 1, 57±69.
Frone, M. R., Russell, M. and Cooper, M. L. (1992a). `Antecedents and outcomes of work±family con¯ict:
testing a model of the work±family interface', Journal of Applied Psychology, 77, 65±78.
Frone, M. R., Russell, M. and Cooper, M. L. (1992b). `Prevalence of work±family con¯ict:
Are work and family boundaries asymmetrically permeable?', Journal of Organizational Behavior, 13,
723±729.
Gaillard, A. W. K. and Wientjes, C. J. E. (1994). `Mental load and work stress as two types of energy
mobilation. Special Issue: A healthier work environment', Work and Stress, 8, 141±152.
Greenberg, E. and Grunberg, L. (1995). `Work alienation and problem alcohol behavior', Journal of Health
and Social Behavior, 36, 83±102.
Greenhalgh, L. (1982). `Maintaining organizational e€ectiveness during organizational retrenchment',
Journal of Applied Behavioral Science, 18, 155±170.
Greenhalgh, L. (1991). `Organizational Coping Strategies'. In: Hartley, J., Jacobson, D., Klandermans, B.
and Van Vuuren, T. (Eds.), pp. 172±199, Job Insecurity. Coping with Jobs at Risk, Sage, London.
Gutek, B. A., Searle, S. and Klepa, L. (1991). `Rational versus gender role explanations for work±family
con¯ict', Journal of Applied Psychology, 76, 560±568.
Hartley, J., Jacobson, D., Klandermans, B. and van Vuuren, T. (Eds.) (1991). Job Insecurity: Coping with
Jobs at Risk, Sage, London.
Higginbottom, S. F., Barling, J. and Kelloway, E. K. (1993). `Linking retirement experiences and marital
satisfaction: a mediational model', Psychology and Aging, 8, 508±516.
Himle, D. P., Jayaratne, S. and Thyness, P. (1989). `The e€ect of emotional support on burnout, work stress
and mental health among Norwegian and American social workers', Journal of Social Service Research,
13, 27±45.
Houtman, I. L. D., Bongers, P. M., Smulders, P. G. W. and Kompier, M. A. J. (1994). `Psychosocial
stressors at work and musculoskeletal problems', Scandinavian Journal of Work, Environment and Health,
20, 139±145.
Hughes, D. and Galinsky, E. (1994). `Work experiences and marital interactions: elaborating the
complexity of work', Journal of Organizational Behavior, 15, 423±438.
Hughes, D., Galinsky, E. and Morris, A. (1992). `The e€ects on job characteristics on marital quality:
specifying linking mechanism', Journal of Marriage and the Family, 54, 31±42.

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
894 S. MAUNO AND U. KINNUNEN

Hurrell, J. J. and Murphy, L. R. (1992). `An overview of occupational stress and health'. In: Rom,
W. M. (Ed.), pp. 675±684, Environmental and Occupational Medicine, Second Edition, Little Brown,
Boston.
Jacobson, D. (1991a). `The conceptual approach to job insecurity', In: Hartley, J. et al. (Eds.), pp. 23±39,
Job Insecurity: Coping with Jobs at Risk, Sage, London.
Jacobson, D. (1991b). `Toward theoretical distinction between the stress components of the job insecurity
and job loss experiences', Research in the Sociology of Organizations, 9, 1±19.
Jones, F. and Fletcher, B. (1993). `An empirical study of occupational stress transmission in working
couples', Human Relations, 46, 881±903.
Jones, F. and Fletcher, B. (1996). `Taking work home: a study of daily ¯uctuations in the work stressors,
e€ects on moods and impacts on marital partners', Journal of Occupational and Organizational
Psychology, 69, 89±106.
Jones-Johnson, G. and Johnson, W. R. (1992). `Subjective underemployment and psychosocial stress: the
role of perceived social and supervisor support', Journal of Social Psychology, 132, 11±21.
JoÈreskog, K. G. and SoÈrbom, D. (1996). LISREL 8: User's Reference Guide, Scienti®c Software Inter-
national, Inc., Chicago.
Karambayaa, R. and Reilly, A. H. (1992). `Dual earner couples: attitudes and actions in restructuring work
for family', Journal of Organizational Behavior, 13, 585±601.
Karasek, R. (1979). `Job demands. Job decision latitude and mental strain: implications for job redesign',
Administrative Science Quarterly, 24, 285±308.
Karasek, R. (1989). `Control in the workplace and its health-related aspects'. In: Sauter, S. L., Hurrell, J. J.
and Cooper, C. L. (Eds.), Job Control and Worker Health, Wiley, Chichester.
Karasek, R. and Theorell, T. (1990). Healthy Work, Basic Books, New York.
Kauppinen-Toropainen, K. (1993). `Women in the labor force in FinlandÐa long history', OECD Panel
Group on Women, Work and Health, National Report: Finland, pp. 1±14, The Ministry of Social A€airs
and Health, The Institute of Occupational Health, Helsinki.
Kauppinen-Toropainen, K., Kandolin, I. and Haavio-Mannila, E. (1988). `Sex segregation of work and the
quality of women's work', Journal of Organizational Behavior, 9, 15±27.
Kelloway, E. K. and Barling, J. (1995). `Stress, control, well-being, and marital functioning: a causal
correlational analysis', In: Keita, G. P. and Hurrell, J. J. (Eds.), pp. 241±251, Job Stress in Changing
Workforce, American Psychological Association, Washington, USA.
Kelly, A. L. and Berthelsen, D. C. (1995). `Preschool teacher's experiences of stress', Teaching and Teacher
Education, 11, 345±357.
Kinnunen, U. and NaÈtti, J. (1994). `Job insecurity in Finland', European Work and Organizational
Psychologist, 4, 297±321.
Kuhnert, K. W., Sims, R. R. and Lahey, M. A. (1989). `The relationship between job security and employee
health', Group & Organization Studies, 14, 399±410.
Lai, G. (1995). `Work and family roles and psychological well-being in urban China', Journal of Health and
Social Behavior, 36, 11±37.
Lambert, S. J. (1990). `Processes linking work and family: a critical review and research agenda', Human
Relations, 45, 239±257.
Lambert, S. J. (1991). `The combined e€ects of job and family characteristics on the job satisfaction, job
involvement, and intrinsic motivation of men and women workers', Journal of Organizational Behavior,
12, 341±363.
Larson, J. H., Wilson, S. M. and Beley, R. (1994). `The impact of job insecurity on marital and family
relationship', Family Relations, 43, 138±143.
Leiter, M. P. and Durup, M. J. (1996). `Work, home, and in-between: a longitudinal study of spillover',
Journal of Applied Behavioral Science, 32, 29±47.
LindstroÈm, K. (1992). `Work organization and well-being of Finnish health care personnel. Fifth
US-Finnish Joint Symposium on Occupational Safety and Health: Occupational epidemics on the
1990s', Scandinavian Journal of Work, Environment and Health, 18 (Suppl. 2), 90±93.
MacEwen, K. E. and Barling, J. (1988). `Interrole con¯ict, family support and marital adjustment of
employed mothers: a short-term longitudinal study', Journal of Organizational Behavior, 9, 241±250.
MacEwen, K. E., Barling, J. and Kelloway, E. K. (1992). `E€ect of acute role overload on marital
functioning', Work and Stress, 6, 117±126.
Maslach, C. and Jackson, S. E. (1981). `The measurement of experienced burnout', Journal of Occupational
Behavior, 2, 99±113.

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)
 EFFECTS OF JOB STRESSORS ON MARITAL SATISFACTION 895

Maslach, C. and Jackson, S. E. (1986). Maslach Burnout Inventory Manual, 2nd ed., Consulting
Psychologist Press, Inc., Palo Alto, California.
Matthews, L. S., Conger, R. D. and Wickrama, K. A. S. (1996). `Work±family con¯ict and marital quality:
mediating processes', Social Psychology Quarterly, 59, 62±79.
McGrath, A., Houghton, D. and Reid, N. (1989). `Occupational stress, and teachers in Northern Ireland',
Work & Stress, 3, 359±368.
McKnight, J. D. and Glass, D. C. (1995). `Perceptions of control, burnout, and depressive symptomat-
ology: a replication and extension', Journal of Consulting and Clinical Psychology, 63, 490±494.
Murphy, L. R. (1995). `Occupational stress management: current status and future directions'. In: Cooper,
C. L. and Rousseau, D. M. (Eds.), Trends in Organizational Behavior, Vol. 2, pp. 1±14, John Wiley &
Sons, Chichester.
Pittman, J. E., Solheim, C. A. and Blanchard, D. (1996). `Stress as a driver of the allocation of housework',
Journal of Marriage and the Family, 58, 456±468.
Repetti, R. L. (1989). `E€ects of daily workload on subsequent behavior during marital interaction:
the roles of social withdrawal and spouse support', Journal of Personality and Social Psychology, 57,
651±659.
Richardsen, A. M. and Burke, R. J. (1993). `Occupational stress and work satisfaction among Canadian
women physicians', Psychological Reports, 72(3, Pt 1), 811±821.
Richardsen, A. M. and Burke, R. J. (1991). `Occupational stress and job satisfaction among physicians: sex
di€erences', Social Science and Medicine, 33, 1179±1187.
Rook, K., Dooley, D. and Catalano, R. (1991). `Stress transmission: the e€ects of husbands' job stressors
on the emotional health of their wives', Journal of Marriage and the Family, 53, 165±177.
Roxburgh, S. (1996). `Gender di€erences in work and well-being: e€ects of exposure and vulnerability',
Journal of Health and Social Behavior, 37, 265±277.
Schaefer, J. A. and Moos, R. H. (1993). `Relationship, task and system stressors in the health-care
workplace. Special issue: work stressors in health care and social service settings', Journal of Community
and Applied Social Psychology, 3, 285±298.
Stets, J. E. (1995). `Job autonomy and control over one's spouse: a compensatory process', Journal of
Health and Social Behavior, 36, 244±258.
Stewart, W. and Barling, J. (1996). `Fathers' work experiences e€ect children's behaviors via job-related
a€ect and parenting behaviors', Journal of Organizational Behavior, 17, 221±232.
Thomas, L. T. and Ganster, D. C. (1995). `Impact of family-supportive work variables on work±family
con¯ict and strain: a control perspective', Journal of Applied Psychology, 80, 6±15.
Turnipseed, D. L. (1994). `An analysis of the in¯uence of work environment variables and moderators on
the burnout syndrome', Journal of Applied Social Psychology, 24, 782±800.
Warr, P. (1987). Work, Unemployment and Mental Health, Oxford University Press, New York.
Weeks, W. A. and Nantel, J. (1995). `The e€ects of gender and career stage on job satisfaction and
performance behavior: a case study. Special issue: gender in the workplace', Journal of Social Behavior
and Personality, 10, 273±288.
Westman, M. and Etzion, D. (1995). `Crossover of stress, strain and resources from one spouse to another',
Journal of Organizational Behavior, 16, 169±181.
Widrich, L. and Ortlepp, K. (1994). `The mediating role of job satisfaction in the work stress±marital
interaction relationship', South African Journal of Psychology, 24, 122±130.
Wiersma, U. J. (1990). `Gender di€erences in job attribute preferences: Work±home role con¯ict and job
level as mediating variables', Journal of Occupational Psychology, 63, 231±243.
Windle, M. and Dumenci, L. (1997). `Parental and occupational stress as predictors of depressive
symptoms among dual-income couples: a multilevel approach', Journal of Marriage and the Family, 59,
625±634.
Zedeck, S. and Mosier, K. (1990). `Work in the family and employing organization', American Psycholo-
gist, 45, 240±252.
Zedeck, S., Maslach, C., Mosier, K. and Skitka, L. (1989). `A€ective response to work and quality of
family life: employee and spouse perspectives'. In: Goldsmith, E. B. (Ed.), pp. 135±157, Work and
Family: Theory, Research, and Application, Sage, California.

Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 879±895 (1999)