WORK AND OCCUPATIONSBarnett, Gareis / REDUCED-HOURS EMPLOYMENT
Although reduced-hours work is widely thought to decrease distress, empirical literature relat-
ing absolute number of hours worked to distress outcomes is inconsistent. Perhaps the trade-off
between giving up some aspects of work for more nonwork time is more stressful for some
employees than for others. The authors tested the hypothesis that difficulty of trade-offs is a more
powerful predictor of quality-of-life indicators (i.e., symptoms of anxiety and depression, job-
role quality, and intention to turnover within 1 year) than is number of hours worked per se in a
non-random sample of 141 reduced-hours physicians in dual-earner couples. Results supported
the hypothesis.
Reduced-Hours Employment
The Relationship Between Difficulty
of Trade-Offs and Quality of Life
ROSALIND CHAIT BARNETT
Brandeis University and Harvard University
KAREN C. GAREIS
Brandeis University
At a time marked by an increase in the number of dual-earner couples
(Bond, Galinsky, & Swanberg, 1998) and in the average workweek of
highly educated employees (Jacobs & Gerson, 1997), pressure has been
building in the United States (Schor, 1991), the United Kingdom (Cooper,
1998), and in Europe (Buessing, 1997) for work schedules that are more ame-
nable to the integration of work and family demands, especially among man-
agers and professionals (Barnett & Lundgren, 1998; Carr et al., 1998;
168
Authors'Note: Data for this analysis were gathered under a grant from the Alfred P.
Sloan Foundation to the first author and Lena Lundgren, Ph.D., at Boston University.
We gratefully acknowledge the contributions of our interviewers, Joyce Buni, Maureen
Crowley, Martie Dumas, Connie Festo, Carla Fink, Phyllis Jansen, Rose Marie Pro-
novost, Marilyn Spiegel, and Jayne Wilson. Correspondence about this article should
be addressed to Rosalind C. Barnett at Women's Studies Program, Brandeis Univer-
sity, Mailstop 082, 415 South Street, Waltham, MA 02454-9110; e-mail: rozbarnett@
mediaone.net.
WORK AND OCCUPATIONS, Vol. 27 No. 2, May 2000 168-187
 2000 Sage Publications, Inc.
Epstein, Seron, Oglensky, & Saute, 1998; European Commission, 1994;
Levy, Flynn, & Kellogg, 1998). Whereas several such schedules entail rear-
ranging the standard 35-hour workweek (e.g., flextime, compressed work-
week, night shifts, and weekend work), others entail a reduction in the aver-
age number of hours worked per week (i.e., part-time). Although the creation
of reduced-hours work options is often motivated by the needs of employers
to limit their financial obligations to employees, it is also thought that the
increased flexibility associated with these options will reduce overall levels
of stress on employees, with consequent beneficial health outcomes
(Buessing, 1996, 1997; Klein, Hyde, Essex, & Clark, 1998; Wethington &
Kessler, 1989). Yet, the empirical literature on the relationship between the
absolute number of hours worked and distress outcomes is inconsistent
(Barnett, 1998).
One possible explanation for this inconsistency is the failure to adequately
model the indirect effects of number of hours worked via mediational
processes such as fit (Barnett, Gareis, & Brennan, 1999). Another possible
explanation is the failure to incorporate psychological aspects of reduced-
hours work. One such aspect is difficulty of trade-offs. Especially among
professional and managerial workers, the decision to decrease hours at work
typically means relinquishing certain aspects of their work. Unlike many
nonprofessionals who may simply do less of the same work when they cut
down on their hours, doctors, for example, often have to eliminate such
activities as teaching and research to reduce their hours. In essence, doctors
trade-off certain professional functions for more nonwork time. For some
professionals, these trade-offs may be easy to make; for others, they may be
more difficult.
In this article, we estimated the relationship between difficulty of trade-
offs and three quality-of-life (QOL) outcomes in a nonrandom sample of 141
married reduced-hours physicians in dual-earner couples.
LITERATURE REVIEW
Recent data reveal (a) an increase over time in the length of the average
workweek among professional employees (Jacobs & Gerson, 1997), (b) a
strong preference among professional workers to substantially reduce their
current work hours (Bond et al., 1998; Jacobs & Gerson, 1997), and (c) the
perception that the absence of reduced-hours career options is a major obsta-
cle to professional advancement (Carr et al., 1998). Interestingly, these senti-
ments were expressed by male and female employees alike and were not
dependent on the presence of young children in the home.
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 169
Despite widespread belief that reduced-hours employment will have a
salutary effect on QOL indicators, there is little empirical support for the
hypothesis that a reduced-hours schedule is related positively to such indica-
tors (Herold & Waldron, 1985; Hyde, Klein, Essex, & Clark, 1995; Klein et al.,
1998; Ozer, Barnett, Brennan, & Sperling, 1998; Wethington & Kessler,
1989). Indeed, most comparative studies report that full-time workers experi-
ence better health than their reduced-hours counterparts (see Barnett, 1998,
for a review). For example, in a longitudinal analysis, White married women
in the United States who decreased their labor force participation from full-
time to low part-time or homemaker reported a significant increase in distress
symptoms over a 3-year period (Wethington & Kessler, 1989). Conversely,
those women who increased their labor force participation from homemaker
or low part-time worker to full-time worker reported a significant decrease in
emotional distress. And, in a national probability sample of middle-aged
women in the United States, Herold and Waldron (1985) reported a tendency
for full-time workers to have the best physical health, part-time workers to
have an intermediate level of health, and women who were not in the labor
force to have the poorest health. Similar findings were reported by Verbrugge
(1989). After controlling for a wide variety of variables in addition to social-
role occupancy, she found that low levels of employment were associated
with poorer physical health.
Even if one did not predict positive outcomes of part-time work for work-
ers in general, one would imagine that the benefits of part-time work would
be felt most strongly in those populations under the most acute work/family
stress. Two such populations in the United States have been studied: (a)
women who have returned to work within a year of having given birth (Klein
et al., 1998) and (b) employed married mothers of preschoolers or mothers of
three or more children (Herold & Waldron, 1985). In a sample of women
interviewed 1 year postpartum, number of hours employed was not signifi-
cantly related to any of the study's four mental health indicators (i.e., depres-
sion, anxiety, anger, and self-esteem). The scores of the homemaker, part-
time, and full-time mothers were indistinguishable on these measures. And,
in concluding their study of physical health, Herold and Waldron (1985)
noted,
No advantage for part-time workers relative to full-time workers was observed
even for married mothers of preschoolers or mothers of three or more children,
for whom one could expect full-time employment to create the greatest role
conflicts and time pressures. These results parallel previous findings that part-
time workers may have better psychological health than women who are out of
the labor force, but differences between part-time workers and full-time work-
ers are inconsistent, even for mothers of young children. (p. 411)
170 WORK AND OCCUPATIONS
These findings seem counterintuitive in light of the strong underlying
assumption that long hours are related to increased distress and that therefore
shorter hours should be associated with less distress. However, just as short
hours are not necessarily or universally associated with better outcomes, nei-
ther are long hours necessarily or universally associated with negative ones
(Barnett & Shen, 1997; Ozer et al., 1998). Indeed, several problems with this
literature have been identified and need to be addressed before we have a
clear idea of the relationship between number of hours worked per se and
health quality-of-life indicators.
METHODOLOGICAL AND CONCEPTUAL
PROBLEMS IN THE LITERATURE
The failure to find the expected relationships between work hours and
quality-of-life indicators may be due to several methodological and concep-
tual problems, some of which apply particularly to between-group studies
comparing full- and part-time employees and others to within-group studies
of part-time workers (see Barnett, 1998, for a review). First, definitions of
full-time and part-time vary from study to study; in some studies, full-time is
defined as 35 or more hours per week, following the Bureau of Labor Statis-
tics'(1995) definition; in other studies, the cutoff is 30 or 32 hours per week.
This problem is particularly acute in studies of the professions in which the
standard workweek may exceed 50 hours and reduced hours may actually
exceed the typical 35-hour workweek. Moreover, most studies are cross-
sectional and do not inquire about the duration of the reduced or full-time
schedule. It is therefore possible that employees who are working reduced
hours are doing so as a temporary adaptation to a life crisis rather than as a
stable career choice. In the present study, we define reduced hours as working
no more than the median number of hours worked per week by other physi-
cians of one's own gender, or 50 hours per week for men and 40 hours per
week for women (Yoon, 1997), and as working the reduced schedule for at
least the preceding 3 months. In addition, the physicians had to consider
themselves to be working reduced hours.
Second, most full-time jobs are "good" jobs, whereas most part-time jobs
are "bad" jobs (Tilly, 1992). Compared to full-time jobs, conventional or
old-concept part-time jobs are those that require few skills, are poorly paid,
offer few (if any) fringe benefits, provide few opportunities for advancement,
provide less job security, and are more likely to be routine boring jobs that
require little training and demand little creativity. New-concept good part-
time jobs, in contrast, are viewed as permanent, they have career potential,
their earnings package includes fringe benefits, and their rate of pay is
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 171
prorated relative to that of comparable full-time jobs. These part-time jobs
began to appear in the late 1960s and are typical of reduced-hours options in
the professions. Examples are job sharing, work sharing, and phased retire-
ment. These forms of part-time work resemble full-time jobs more than they
do conventional part-time jobs. The physicians in this study all had good
part-time jobs.
Third, some part-time workers are involuntarily employed less than full-
time, whereas other workers choose to work part-time. According to Negrey
(1993), about three fourths of part-time workers in the United States were
employed part-time by choice. The remaining one fourth were employed
part-time because their hours had been reduced due to slack work or because
they were unable to find full-time employment. All physicians in the present
study had voluntarily reduced their hours.
Fourth, there is the issue of selection. Women and men who select part-
time work may differ from women and men who select full-time work in
ways that relate systematically to their ability to manage stress. For example,
compared to full-time workers, part-time workers are, on average, younger or
older, less educated, female, married, and if married, married to a full-time
employed partner. In the present study, we limited the sample to married phy-
sicians 30 to 65 years of age and controlled for gender.
Finally, the focus on number of hours per se ignores several psychological
factors. For example, not all part-time workers are employed at their pre-
ferred number of hours. In a 1991 study of Canadian employees, Kahn and
Lang (1992) found that 48% reported that they were working too few hours,
16% were working too many hours, and 36% were working optimal hours. In
the present study, we inquired into the discrepancy between the physician's
actual and preferred work hours and used it as a control in our analyses.
TRADE-OFFS
In addition to the limitations discussed earlier, which we address in the
present analysis, there is another conceptual problem that has not been noted
previously. In most cases, physicians (and other professionals) who decide to
reduce their work hours have to give up certain professional activities.
1
Thus,
few part-time doctors spend much time in such pursuits as research and
teaching. Reduced-hours physicians apparently trade away these activities
for more nonwork time. In a preliminary qualitative study of 24 reduced-
hours physicians conducted on this sample (Lundgren & Barnett, 1997), the
overwhelming majority reported having to give up some professional respon-
sibilities when they decreased their work hours. In contrast, many nonprofes-
sional employees who reduce their hours merely do fewer of the same tasks.
172 WORK AND OCCUPATIONS
For example, a sales clerk who cuts down on work hours would simply spend
fewer hours behind the sales counter.
The idea that we commonly make trade-offs--that is, that we often have to
choose to do one thing rather than another--is not new. For example, the
realization that long work hours require that one trade-off time at home for
time at work has been conceptualized as one component of work-family con-
flict (Frone, Russell, & Cooper, 1992, 1997; Gutek, Searle, & Klepa, 1991;
Hughes, Galinsky, & Morris, 1992; Netemeyer, Boles, & McMurrian, 1996).
More generally, the notion that the demands of one role may preclude the
adequate performance of a second role has been discussed in terms of inter-
role conflict. Indeed, several studies on nonstandard work schedules in the
United States and in Europe (Buessing, 1996, 1997; European Commission,
1994) have focused on interrole conflict, namely on the nonwork (i.e., family
and leisure) implications of various work schedules. For example, Buessing
(1997) in a study of 482 German nurses found that family- and leisure-related
arguments dominated the pros and cons nurses gave regarding five major
nonstandard working time schedules including part-time. Moreover, these
findings were largely comparable to those from a representative German
sample of 2,577 employed persons. Furthermore, these various forms of
work-family conflict have been associated with such QOL outcomes as
depression, psychological distress, life satisfaction, and marital-role quality
(Adams, King, & King, 1996; Frone et al., 1997; Greenhaus & Parasuraman,
1994; Netemeyer et al., 1996).
In contrast, almost no research in the United States or Europe has explored
a second type of conflict that can coexist with interrole conflict: namely,
intrarole conflict. We conceptualize trade-offs as a type of intrarole conflict
in which the incumbent has to relinquish some aspects of the work role
because they cannot all be performed in the reduced time now allocated to
work. However, the concept of trade-offs is not adequately captured by a
focus on conflict. Some physicians who give up particular professional
activities may experience difficulty, whereas others may not. In the present
analysis, we conceptualize difficulty of trade-offs as a continuum ranging
from not at all to very distressing and estimate the relationship between this
variable and three QOL indicators. This conceptualization may help clarify
several previous findings indicating that long work hours are related to better
mental health in certain samples (Barnett & Shen, 1997; Ozer et al., 1998).
The concept of trade-offs may also have relevance for full-time employ-
ees, in particular specialists who advance to administrative or managerial
positions. Such specialist-to-generalist transitions entail similar trade-offs of
certain work responsibilities for others, and successful transitions may be
predicted by difficulty of trade-offs. There is some literature on this aspect of
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 173
career transitions among full-time employees (Dalton, 1989; Hill, 1992;
Schein, 1978); however, most of it is descriptive in nature.
Among full-time employed physicians, the new realities of practicing
medicine in the era of health maintenance organizations (HMOs) are creating
changes that are reshaping the profession and forcing doctors to make trade-
offs previously unimagined. As described by Hoff and McCaffrey (1996),
doctors who maintain a solo or group practice now have to "become more
conscientious business owners" (p. 177) who have to accept an administrator-
manager role and develop management-oriented beliefs to stay in business.
On the other hand, physicians who work for HMOs trade-off a guaranteed
income for less control and independence. This loss of control results in large
part from the increased role now played by third parties.
Many doctors who try to maintain the ideal of the solo practitioner experi-
ence the loss of an "important part of their professional freedom" (Hoff &
McCaffrey, 1996, p. 176) as well as "extreme anger and resentment at the
organizational and financial changes occurring within the health care mar-
ket" (p. 175). Although they chose to maintain their independence and avoid
one type of trade-off, they experienced another type of trade-off: They had to
take on new financial and administrative responsibilities that were associated
with distress. In contrast, doctors employed by HMOs experienced some dis-
tress over the rules of the game they were now expected to follow. For exam-
ple, they could only prescribe medications on the HMO's approved list. Some
doctors experienced this constraint as a trade-off between their particular
practice styles and those mandated by their employers.
More generally, professional career development often entails a series of
transitions marked by taking on new responsibilities and giving up others. If
the new responsibilities are compatible with one's career anchors--that is,
with one's self-perceived talents and abilities, motives and needs, and atti-
tudes and values (Schein, 1978)--then progress will be far smoother than if
they are not.
The individual-level reactions of professionals to changing work condi-
tions that were reported by Hoff and McCaffrey (1996) and that are extended
in the present study help flesh out our understanding of the personal ramifica-
tions of changes that are occurring as professions undergo transformations.
POTENTIAL MODERATORS
Although there is no previous literature to draw on, it is reasonable to
believe that in this sample, the relationship between difficulty of trade-offs
and QOL indicators may vary by age, gender, and medical specialization.
Conceivably, older physicians who have taught and done research for several
174 WORK AND OCCUPATIONS
years might be less disturbed about giving up those activities than their
younger counterparts. Also, certain widely held gender theories, although
challenged in the empirical literature, suggest that the relationship between
trade-offs and QOL indicators may be stronger among men than among
women. Specifically, difficulty of trade-offs might be more strongly related
to QOL indicators for men than for women because the work role and work
experiences are thought to be more central to men's core identity than to
women's (Barnett, Carr, et al., 1998). Finally, certain medical specialties may
more easily accommodate reduced-hours schedules than others (Barnett &
Lundgren, 1998), thereby either attenuating or amplifying the relationship
between trade-offs and QOL indicators.
NEGATIVE AFFECTIVITY
Serious questions have been raised about the common practice of relying
on self-report measures to assess the relationship between subjective role
experiences on one hand and QOL indicators on the other (Burke, Brief, &
George, 1993). Such self-report measures may reflect a common underlying
dimension of negative affectivity (NA), thereby leading to spurious relation-
ships, particularly in cross-sectional analyses. Following recent caveats in
the literature (Brennan & Barnett, 1998), we include trait anxiety (Spiel-
berger, 1983), a recognized measure of negative affectivity, as a control
variable.
In sum, in a sample of married physicians who are working reduced hours
and whose spouses are employed, we test the hypothesis that difficulty of
trade-offs will be significantly associated with three stress-related QOL out-
comes. Specifically, we test this hypothesis by estimating the relationship
between difficulty of trade-offs on one hand and intention to leave one's cur-
rent job in 1 year, current job-role quality, and self-reported symptoms of
anxiety and depression (together comprising an index of psychological dis-
tress) on the other. We also estimate the two-way interaction effects of gen-
der, age, and medical specialization on these relationships.
METHOD
SAMPLE
One hundred and forty-one Boston-area physicians who were employed
on a reduced-hours schedule (116 women and 25 men) and their employed
spouses were interviewed as part of a larger study of the anticipated and
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 175
unanticipated consequences of reduced-hours work on physicians, their
spouses, and their employing organizations (Barnett et al., 1999; Hart-
well, Barnett, Borgatti, & Lundgren, 1999; Lundgren, Gareis, Fleisher-
Cooperman, & Fitzgerald, 1998).
Reduced hours was defined as (a) working no more than the median
number of hours worked per week by other physicians of one's own gender,
or 50 hours per week for men and 40 hours per week for women (Yoon,
1997), for at least the preceding 3 months; and (b) considering oneself to be
working reduced hours. Although both physicians and spouses were inter-
viewed, only the data from the 141 physicians were used in the present
analyses.
We used several strategies to develop the sample including random sam-
pling of area part-time physicians from the registry of the Massachusetts
Board of Certification in Medicine, requesting nominations from project par-
ticipants, and contacting every area HMO, hospital, and practice partnership
and inviting their participation either by nominating eligible staff physicians
or by publicizing the study and asking interested physicians to contact us.
Because the registry data did not include total work hours, we used patient
care hours as a proxy selection variable.
The conservatively calculated response rate was 35.5% among the ran-
domly sampled physicians and 43.0% among the nominated physicians;
these response rates compare favorably with those reported in previous stud-
ies of physicians (Parsons, Warnecke, Czaja, Barnsley, & Kaluzny, 1994;
Sobal et al., 1990).
2
PROCEDURES
Each physician was interviewed separately by a trained interviewer. The
interviews took approximately 1 hour and were conducted at a time and place
convenient to the participant. Prior to the interview, each participant received
a mailed survey that took about 20 minutes to complete and was collected at
the time of the interview. The interviews covered the quality of the partici-
pants'major social roles--partner, employee, and where appropriate, parent--
along with QOL indicators. The measures of interest were administered dur-
ing the face-to-face data collection.
MEASURES
Trade-offs. As an initial effort to operationalize difficulty of trade-offs, we
asked the physicians to indicate on a 5-point scale from 1 (none) to 5 (a great
deal) the degree of distress experienced due to the discrepancy between the
176 WORK AND OCCUPATIONS
professional activities they would like to perform and their current work
arrangement. This item was preceded by one asking participants to estimate
the percentage of time they were currently spending in each of five profes-
sional activities (i.e., patient care, research, teaching, administration, and
other) and a second item asking them to indicate the degree to which they
were satisfied with the percentage of time they were currently spending on
each professional activity.
Anxiety and depression (i.e., psychological distress). Respondents indi-
cated on a 5-point scale from 0 (not at all) to 4 (extremely) how often in the
past week they were bothered by each of 10 symptoms of anxiety and 14
symptoms of depression (Derogatis, 1975). Anxiety items include "feeling
fearful" and "trembling"; depression items include "feeling blue" and "poor
appetite." Anxiety and depression scores were combined to create a measure
of psychological distress (Barnett, Marshall, Raudenbush, & Brennan,
1993). The combined score has excellent internal consistency, with Cron-
bach's alphas of .90 in a previous study (Barnett, Marshall, & Singer, 1992)
and .91 in the present sample. These figures are similar to those reported by
Derogatis (1983).
Turnover was operationalized as a single item asking participants to indi-
cate the likelihood that they will voluntarily terminate employment with their
organization within the next 12 months using a 7-point scale from 1
(extremely unlikely) to 7 (extremely likely).
Job-role quality was assessed by a 28-item measure on which respondents
were instructed to think about their job as it is right now and to indicate on a
4-point scale from 1 (not at all) to 4 (extremely) to what extent each of the
items was currently rewarding or of concern (Barnett & Brennan, 1995,
1997). The items covered job conditions in areas such as skill discretion,
decision authority, schedule control, job demands, pay adequacy, job secu-
rity, and relations with supervisor. Concern items were negatively weighted
and reward items positively weighted in constructing the role-quality score,
which was the weighted average of the item scores (Barnett et al., 1993).
Internal consistency is good, with Cronbach's alphas of .79 for rewards and
.83 for concerns in the present sample.
Work hours was assessed by asking respondents to estimate the number of
hours worked in an average workweek.
Age is self-explanatory.
Gender was coded as a dummy variable (1 = female, 0 = male).
Medical specialization was operationalized as a dummy variable (1 =
medical specialty, 0 = else).
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 177
Household income per capita was calculated by dividing household
income by the number of children in the home plus two. Household income
here is the sum of the two partners' salaries plus their unearned income.
Because the distribution of this variable is highly skewed, we used the natural
log of per capita income.
Number of children younger than 18 years of age living at home is self-
explanatory.
Negative affectivity was assessed by the Trait Anxiety Scale (Spielberger,
1983), a 10-item frequency of feelings scale. Test-retest correlations ranged
from .73 to .86 in college populations over a 2-year period (Spielberger,
1983) and was .77 over a 1-year period in a sample of full-time employed
dual-earner couples (Barnett, Brennan, Raudenbush, Pleck, & Marshall,
1995).
Preferred work hours was assessed by asking respondents to rate their cur-
rent hours compared to those they would ideally like to work as (1) "too few,"
(2) "roughly the right number," and (3) "too many." For purposes of these
analyses, responses were dichotomized into a dummy variable (1 = ideal, 0 =
not ideal).
RESULTS
DESCRIPTIVE RESULTS
The majority of the sample was White (90.0%), followed by Hispanic
(4.3%), Asian (2.9%), and African American (0.7%); 2.1% described them-
selves as other. On average, the physicians were 42.1 years of age (men:
mean = 44.5, SD = 9.8; women: mean = 41.6, SD = 5.6), had been working
their current reduced-hours schedules for about 4 years (mean = 47.4 months,
SD = 47.3), and were parents (95.7%) with 2.1 children (SD = 0.9). On aver-
age, the male physicians worked 33.5 hours per week (SD = 10.4, range =
10-45) compared to 29.4 hours (SD = 8.7, range = 1-50) for the female physi-
cians.
3
Although these physicians voluntarily reduced their hours, only
53.9% reported working their preferred number of hours, 37.6% reported
working more, and 8.5% reporting working fewer.
The sample was largely female (82.3% vs. 17.7%), reflecting the relative
percentages of male and female doctors in the population. The major reason
given by the women who declined to participate in the study was that their
husbands, typically full-time employed doctors or other professionals, were
unwilling. Thus, the sample may be biased toward women doctors whose
husbands are not full-time employed doctors or other professionals or,
178 WORK AND OCCUPATIONS
alternatively, are more willing than their counterparts to devote the time nec-
essary to complete the study's requirements. We have no way of ascertaining
how this bias might have affected the results.
Overall, the physicians reported moderate difficulty of trade-offs, with an
average rating of 2.7 (SD = 0.9) on a scale from 1 to 5 (men: mean = 2.4, SD =
1.0; women: mean = 2.8, SD = 0.9).
The intercorrelations among the study variables are shown in Table 1. As
can be seen, the three outcome variables were moderately correlated and in
the predicted direction (from r = ­.39 to r = .17), and difficulty of trade-offs
was moderately correlated in the expected direction with each of the QOL
indicators (from r = ­.54 to r = .25), providing some assessment of the con-
struct validity of this single-item measure.
HYPOTHESIS TESTING
We estimated a series of main effects simultaneous regression models
with difficulty of trade-offs as the predictor; number of hours worked, pre-
ferred work hours, age, gender, per capita household income, number of chil-
dren living at home, and medical specialization as controls; and one of the
three QOL indicators (i.e., psychological distress, turnover, and job-role
quality) as the outcome. In addition, we tested whether these relationships
were moderated by age, gender, and medical specialization. To that end, for
each regression model, we entered three interaction terms of the form Diffi-
culty of Trade-Offs × Moderator.
Each main effects regression model was significant, and in each model,
difficulty of trade-offs was a significant predictor (see Table 2). Specifically,
as hypothesized, difficulty of trade-offs was positively associated with psy-
chological distress and negatively associated with both job-role quality and
intention to turnover. Moreover, after taking into account difficulty of trade-
offs (and the control variables), number of hours worked per week on average
was not significantly related to any of the three QOL outcomes. Furthermore,
examination of the standardized regression coefficients indicates that in each
model, difficulty of trade-offs accounted for approximately three to eight
times the percentage of variance accounted for by work hours. Finally, the
inclusion of the set of interaction terms did not result in a significant incre-
ment to R
2
over and above that associated with the main effects models. Thus,
the relationships between difficulty of trade-offs and distress, turnover, and
job-role quality did not differ by age, gender,
4
or medical specialization.
To control for the possible confounding effect of negative affectivity, we
repeated the analyses including NA as an additional control variable. The
results (not shown) were virtually unchanged. Thus, the relationships
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 179
TABLE1:IntercorrelationsAmongMeasures
HOURSPREFAGESEXINCKIDSMEDSPDISTRJOBTURN
TRADE.13.07­.17*.15­.11­.06.03.38***­.54***.25**
HOURS.33***­.26**­.17*­.06.06­.13.01­.24**.00
PREF­.22**.16­.03.20*­.22**.06­.17*­.01
AGE­.17*.27**­.21*.05­.11*.14­.05
SEX.01.16­.03.14­.01­.17*
INC­.46***­.04­.06.05­.10
KIDS­.04­.03.00­.09
MEDSP.10­.04.04
DISTR­.26**.17*
JOB­.39***
NOTE:N=141.IntercorrelationsamongTRADE(difficultyoftrade-offs),HOURS(workhours),PREF(preferredworkhours),AGE,SEX(0=
male,1=female),INC(percapitahouseholdincome),KIDS(numberofchildrenathome),MEDSP(medicalspecialty;0=other,1=medi-
cal),DISTR(psychologicaldistress),JOB(jobrole-quality),andTURN(intentiontoturnover).
*p<.05.**p<.01.***p<.001.
180
TABLE2:RelationshipsBetweenDifficultyofTrade-OffsandThreeQuality-of-Life(QOL)Indices
PsychologicalDistressJob-RoleQualityIntentiontoTurnover
VariableBbSEBbSEBbSE
Difficultyoftrade-offs4.04***.351.02­0.42***­.570.060.62**.300.19
Workhours­0.09­.070.11­0.01­.070.01­0.02­.110.02
Preferredhours­2.16­.101.820.10.080.10­0.06­.020.33
Age­0.05­.030.15­0.00­.000.010.00.010.03
Gender0.35.012.550.18.100.15­1.30**­.260.47
Householdincome­1.35­.034.76­0.15­.050.27­1.31­.150.87
Childrenathome­0.71­.061.16­0.01­.010.07­0.11­.050.21
Medicalspecialty2.24.111.77­0.08­.060.10­0.01­.000.33
ModelR2
.17.35.16
AdjustedR2
.12.31.10
NOTE:N=141.Genderiscodedas0=male,1=female.
**p<.01.***p<.001.
181
between difficulty of trade-offs and three QOL indicators were not affected
by the tendency toward experiencing the world negatively.
DISCUSSION
The main finding of this study of a nonrandom sample of 141 Boston-area
physicians who were married to employed spouses and who were employed
on a reduced-hours schedule is that difficulty of trade-offs is a much stronger
predictor of three QOL indicators (i.e., current job-role quality, intention to
turnover, and psychological distress) than is number of hours worked. Thus,
the subjective meaning of reducing work hours has to be taken into account in
assessing the QOL correlates of reduced-hours career options in the profes-
sions. Moreover, the relationship between difficulty of trade-offs and QOL
indicators did not differ by age, gender, or medical specialization. It is also
possible that the limited range of actual work hours affected our ability to
detect a significant relationship between work hours and QOL indicators.
However, previous research with full-time employed men and women also
failed to find such a relationship (see Barnett, 1998, for a review). Finally,
given the specialized nature of the sample, caution should be exercised in
generalizing the findings.
CONCLUSIONS
These findings extend previous research comparing the predictive validity
of subjective and objective indicators of job experiences. The general con-
sensus in this literature is that subjective indicators are stronger predictors of
distress outcomes than are objective indicators. For example, several U.S.
studies indicate that subjective job-role quality is a better predictor of psy-
chological distress than is number of hours worked or occupational prestige
(Barnett, 1998; Hyde et al., 1995; Klein et al., 1998). In contrast to earlier
studies that focused on overall job-role quality or on subjective experiences
with respect to specific job conditions (Barnett & Brennan, 1995, 1997), the
focus on trade-offs draws attention to the experience of intrarole conflict. In
an effort to gain flexibility and increased time for nonwork commitments,
many professionals may not fully appreciate how disquieting it is to relin-
quish certain aspects of their professional lives. According to Schein (1978),
the subjective importance of particular aspects of work may not be visible
"until one faces choices" (p. 129). Alternatively, however, it may be that
182 WORK AND OCCUPATIONS
physicians who are less interested in such activities as research and teaching
are especially drawn to part-time work.
Results from this initial study suggest that the concept of difficulty of
trade-offs has predictive validity and warrants further study. One line of
future research would be to further develop the measure. A revised measure
ought to reflect the fact that for some professionals, reducing work hours may
be beneficial in that it allows them to relinquish aspects of their work that they
experience as onerous (S. MacDermid, personal communication, September
24, 1998). One would also like to assess how satisfying or distressing were
the actual activities that were given up. It is also likely that reduced-hours
workers vary in the extent to which they were actually required to relinquish
professional activities. Items assessing these various aspects of trade-offs
should be included in any future development of this measure. Thus, to fur-
ther advance the trade-offs construct, additional elaboration of the measure
needs to be done.
Future research is also needed to determine whether other subjective indi-
cators might also be important to study. For example, many researchers have
suggested that there are special rewards and concerns associated with part-
time compared to full-time work. Major among the rewards often mentioned
are: having more time for yourself, your children, and your partner; feeling
less exhausted at work; and being better able to integrate the demands of work
and those of family. Often-mentioned concerns include: lower pay, less job
security, and lower probability of being promoted. A measure of the degree to
which such aspects of part-time work were experienced as rewarding or of
concern might be another fruitful indicator to examine.
Although this analysis of trade-offs was limited to reduced-hours workers
in the medical profession, it seems plausible that the findings would general-
ize to other professionals and to certain full-time employees who may also
experience intrarole conflict as a result of having to trade away valued aspects
of their professional work for others. Among reduced-hours lawyers, work-
ing on high-profile cases and being on a partnership track are traded off for
more nonwork time (Epstein et al., 1998). And, accountants working reduced
hours trade the development of in-house relationships, which are often key to
partnership, for more nonwork time (Levy et al., 1998). Moreover, several
studies indicate that full-time specialists in industry or the academy who
move into management or administration experience considerable intrarole
conflict. They too must trade-off certain professional activities for others,
and it seems likely that difficulty of trade-offs for these employees might pre-
dict QOL outcomes just as it does for part-time employees who have made
trade-offs to reduce their hours.
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 183
For example, in a richly detailed study of the early career transition of 19
new sales and marketing managers, Hill (1992) suggested that the change
from being an individual producer, a doer, to being a manager requires giving
up certain professional activities, especially those associated with prior suc-
cess and self-esteem, and acquiring new less familiar and less valued activi-
ties. New managers have to trade time in direct performance of technical
work for time in people management (i.e., supervising and coordinating oth-
ers'efforts), building a cooperative relationship with the boss, network build-
ing, and managing office politics. Perhaps because of their discomfort over
the trade-offs they have to make, "many new managers never adjust success-
fully to managerial responsibilities" (p. 2) in spite of their established qualifi-
cations as individual contributors.
Finally, previous research by Hoff and McCaffrey (1996) indicated that
the emergence of HMOs has forced many full-time physicians to make intra-
role trade-offs that are often experienced as distressing. Taken together, these
findings indicate that the trade-offs concept may well be applicable to a range
of occupational transitions, some of which are entered into voluntarily, others
of which are imposed by external circumstances.
NOTES
1. Demographic data suggest that physicians who plan to work reduced hours select particu-
lar areas of medical specialization and not others. For example, more than half of the reduced-
hours physicians were medical specialists as opposed to surgeons or general practitioners. Of
those with medical specialties, the largest single group, accounting for approximately 44%, was
psychiatrists. This percentage is roughly eight times as high as the comparable percentage
(5.4%) nationally (American Medical Association, 1997).
2. A response rate cannot be calculated for those physicians (n = 22) who contacted us. How-
ever, among the randomly sampled physicians (n = 82), the completion rate (interviews divided
by the sum of interviews and refusals) was 49.1%, whereas the more conservatively calculated
response rate for potentially eligible participants (interviews divided by the sum of interviews,
refusals, and estimated eligibles among uncontactable respondents) was 35.5%. The compara-
ble figures for the nominated sample (n = 40) were 78.4% and 43.0%. Previous studies of physi-
cians report response rates from about 60% to 65%. However, these studies, which usually in-
volve much shorter surveys than the present study (Parsons, Warnecke, Czaja, Barnsley, &
Kaluzny, 1994; Sobal et al., 1990), required only the physician to agree to participate rather than
requiring the spouse's agreement as well. The majority of the physicians in our sample (60.3%)
were married to other physicians, most employed full-time, with the same notoriously high re-
fusal rates as found in most physician studies. The probability that two spouses will agree to par-
ticipate, assuming independent participation likelihoods ranging from .60 to .65, is only .36 to
.42. These figures approximate the conservatively calculated overall response rates achieved in
this study, with cooperation rates substantially higher.
3. Close examination of the survey protocols indicated that 9 female physicians who reported
during the screening that they worked 40 or fewer hours per week and considered themselves to
184 WORK AND OCCUPATIONS
be working reduced hours indicated during the interview that they were working more than 40
hours yet still considered themselves to be working reduced hours. We retained them in the sam-
ple after rerunning the analyses with these cases excluded and determining that the results were
not affected by their inclusion.
4. It is important to note that the relatively small number of male physicians in our sample
may have reduced our ability to detect any but large gender effects.
REFERENCES
Adams, G. A., King, L. A., & King, D. W. (1996). Relationships of job and family involvement,
family social support, and work-family conflict with job and life satisfaction. Journal of Ap-
plied Psychology, 81, 411-420.
American Medical Association. (1997). Physician characteristics and distribution in the U.S.,
1996-1997. Chicago: Author.
Barnett, R. C. (1998). Toward a review and reconceptualization of the work/family literature.
Genetic, Social and General Psychology Monographs, 124, 125-182.
Barnett, R. C., & Brennan, R. T. (1995). The relationship between job experiences and psycho-
logical distress: A structural equation approach. Journal of Organizational Behavior, 16,
259-276.
Barnett, R. C., & Brennan, R. T. (1997). Change in job conditions, change in psychological dis-
tress, and gender: A longitudinal study of dual-earner couples. Journal of Organizational
Behavior, 18, 253-274.
Barnett, R. C., Brennan, R. T., Raudenbush, S. W., Pleck, J. H., & Marshall, N. L. (1995). Change
in job and marital experiences and change in psychological distress: A longitudinal study of
dual-earner couples. Journal of Personality and Social Psychology, 69, 839-850.
Barnett, R. C., Carr, P., Boisnier, A. D., Ash, A., Friedman, R. H., Moskowitz, M. A., & Szalacha, L.
(1998). Relationship of gender and career motivation to medical faculty members' produc-
tion of academic publications. Academic Medicine, 76, 180-186.
Barnett, R. C., Gareis, K. C., & Brennan, R. T. (1999). Fit as a mediator of the relationship be-
tween work hours and burnout. Journal of Occupational Health Psychology, 4, 307-317.
Barnett, R. C., & Lundgren, L. (1998). Dual-earner couples and the decision to work less: A con-
ceptual model. Community, Work and Family, 1, 273-295.
Barnett, R. C., Marshall, N. L., Raudenbush, S., & Brennan, R. (1993). Gender and the relation-
ship between job experiences and psychological distress: A study of dual-earner couples.
Journal of Personality and Social Psychology, 64, 794-806.
Barnett, R. C., Marshall, N. L., & Singer, J. D. (1992). Job experiences over time, multiple roles,
and women's mental health: A longitudinal study. Journal of Personality and Social Psy-
chology, 62, 634-644.
Barnett, R. C., & Shen, Y.-C. (1997). Gender, high- and low-schedule-control housework tasks,
and psychological distress. Journal of Family Issues, 18, 403-428.
Bond, J. T., Galinsky, E., & Swanberg, J. E. (1998). The 1997 national study of the changing
workforce. New York: Families and Work Institute.
Brennan, R. T., & Barnett, R. C. (1998). Negative affectivity: How serious a threat to self-report
studies of psychological distress. Women's Health: Research on Gender, Behavior, and Pol-
icy, 4, 369-384.
Buessing, A. (1996). Social tolerance of working time scheduling in nursing. Work and Stress,
10, 238-250.
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 185
Buessing, A. (1997). Working time scheduling and the relation between work, family and lei-
sure. In G. Bosch, D. Meulders, & F. Michon (Eds.), Working time: New issues, new norms,
new measures (pp. 39-60). Brussels, Belgium: Ed. du Dulbea.
Bureau of Labor Statistics. (1995). Labor force statistics: Annual average data 1994. Washing-
ton, DC: Author.
Burke, M. J., Brief, A. P., & George, J. M. (1993). The role of negative affectivity in understand-
ing relations between self-reports of stressors and strains: A comment on the applied psy-
chology literature. Journal of Applied Psychology, 78, 402-412.
Carr, P. L., Ash, A. S., Friedman, R. H., Scaramucci, A., Barnett, R. C., Szalacha, L., Palepu, A., &
Moskowitz, M. A. (1998). The relationship of family responsibilities and gender to the
productivity and career satisfaction of medical faculty. Annals of Internal Medicine, 129,
532-538.
Cooper, C. (1998). The changing nature of work. Community, Work and Family, 1, 313-317.
Dalton, G. (1989). Developmental view of careers in organizations. In M. B. Arthur & D. T. Hall
(Eds.), Handbook of career theory (pp. 89-109). New York: Cambridge University Press.
Derogatis, L. R. (1975). The SCL-90-R. Baltimore: Clinical Psychometrics.
Derogatis, L. R. (1983). SCL-90-R: Administration, scoring, and procedures manual for the re-
vised version and other instruments of the psychopathology rating scale series. Towson,
MD: Clinical Psychometric Research.
Epstein, C. F., Seron, C., Oglensky, B., & Saute, R. (1998). The part-time paradox: Time norms,
professional life, family and gender. Boston: Routledge Kegan Paul.
European Commission. (1994). Employment in Europe. Brussels, Belgium: Author.
Frone, M. R., Russell, M., & Cooper, L. M. (1992). Prevalence of work-family conflict: Are
work and family boundaries asymmetrically permeable? Journal of Organizational Behav-
ior, 13, 723-729.
Frone, M. R., Russell, M., & Cooper, L. M. (1997). Relation of work-family conflict to health
outcomes: A four-year longitudinal study of employed parents. Journal of Occupational and
Organizational Psychology, 70, 325-335.
Greenhaus, J. H., & Parasuraman, S. (1994). Work-family conflict, social support and well-
being. In M. J. Davidson & R. J. Burke (Eds.), Women in management (pp. 213-229). Lon-
don: Paul Chapman Publishing, LTD.
Gutek, B. A., Searle, S., & Klepa, L. (1991). Rational versus gender role explanations for work-
family conflict. Journal of Applied Psychology, 76, 560-568.
Hartwell, J. K., Barnett, R. C., Borgatti, S., & Lundgren, L. (1999, August). When medical man-
agers'needs encounter needs of reduced-hour physicians: Two organizational subcultures
confront change. Paper presented at the annual meeting of the Academy of Management,
Chicago.
Herold, J., & Waldron, I. (1985). Part time employment and women's health. Journal of Occupa-
tional Medicine, 27, 405-412.
Hill, L. A. (1992). Becoming a manager: How new managers master the challenges of leader-
ship. New York: Penguin.
Hoff, T. J., & McCaffrey, D. P. (1996). Adapting, resisting, and negotiating: How physicians
cope with organizational and economic change. Work and Occupations, 23, 165-189.
Hughes, D., Galinsky, E., & Morris, A. (1992). The effects of job characteristics on marital qual-
ity: Specifying linking mechanisms. Journal of Marriage and the Family, 54, 31-42.
Hyde, J. S., Klein, M. H., Essex, M. J., & Clark, R. (1995). Maternity leave and women's mental
health. Psychology of Women Quarterly, 19, 257-285.
186 WORK AND OCCUPATIONS
Jacobs, J. A., & Gerson, K. (1997). The endless day or flexible office? Working hours, work-
family conflict, and gender equity in the modern workplace (report to Alfred P. Sloan Foun-
dation). New York.
Kahn, S., & Lang, K. (1992). Constraints on the choice of work hours: Agency versus specific
capital. Journal of Human Resources, 27, 661-678.
Klein, M. H., Hyde, J. S., Essex, M. J., & Clark, R. (1998). Maternity leave, role quality, work in-
volvement, and mental health one year after delivery. Psychology of Women Quarterly, 22,
239-266.
Levy, E. S., Flynn, P. M., & Kellogg, D. M. (1998). Customized work arrangements in the ac-
counting profession: An uncertain future. New York: Alfred P. Sloan Foundation.
Lundgren, L., & Barnett, R. C. (1997, July). Dual-career families and reduced-hours career
paths in the medical profession. Paper presented at the annual conference on work-family
roles, Louisville, Kentucky.
Lundgren, L., Gareis, K. C., Fleisher-Cooperman, J., & Fitzgerald, T. (1998, November).
Choosing a reduced-hours schedule: Factors influencing this decision in two-earner cou-
ples. Paper presented at Work and Family: Today's Realities and Tomorrow's Visions, a con-
ference sponsored by the Alfred P. Sloan Foundation, the Business and Professional
Women's Foundation, and the Wellesley College Center for Research on Women, Boston.
Negrey, C. (1993). Gender, time, and reduced work. Albany, NY: State University of New York
Press.
Netemeyer, R. G., Boles, J. S., & McMurrian, R. (1996). Development and validation of
Work-Family Conflict and Family-Work Conflict scales. Journal of Applied Psychology,
81, 400-410.
Ozer, E. M., Barnett, R. C., Brennan, R. T., & Sperling, J. (1998). Does childcare involvement in-
crease or decrease distress among dual-earner couples? Women's Health: Research on Gen-
der, Behavior, and Policy, 4, 285-311.
Parsons, J. A., Warnecke, R. B., Czaja, R. F., Barnsley, J., & Kaluzny, A. (1994). Factors associ-
ated with response rates in a national survey of primary care physicians. Evaluation Review,
18, 756-766.
Schein, E. H. (1978). Career dynamics: Matching individual and organizational needs. Read-
ing, MA: Addison-Wesley.
Schor, J. B. (1991). The overworked American: The unexpected decline of leisure. New York:
Basic Books.
Sobal, J., DeForge, B. R., Ferentz, K. S., Muncie, H. L., Valente, C. M., & Levine, D. M. (1990).
Physician responses to multiple questionnaire mailings. Evaluation Review, 18, 711-722.
Spielberger, C. D. (1983). Manual for the State-Trait Anxiety Inventory (Form Y ed.). Palo Alto,
CA: Consulting Psychologist Press.
Tilly, C. (1992). Two faces of part time work: Good and bad part-time jobs in U.S. service indus-
tries. In B. D. Warme, K.L.P. Lundy, & L. A. Lundy (Eds.), Working part-time: Risks and
opportunities. New York: Praeger.
Verbrugge, L. M. (1989). The twain meet: Empirical explanations of sex differences in health
and mortality. Journal of Health and Social Behavior, 30, 282-304.
Wethington, E., & Kessler, R. C. (1989). Employment, parental responsibility, and psychologi-
cal distress. Journal of Family Issues, 10, 527-546.
Yoon, Y. (1997). Reduced-hours careers in medicine. Unpublished manuscript.
Barnett, Gareis / REDUCED-HOURS EMPLOYMENT 187