Men's
and
Women's Definitions of
"Good"
Jobs
Similarities
and
Differences by
Age and Across
Time
PAMELA
S.TOLBERT
PHYLLIS
MOEN
Cornell University
Abstract:
Whether and
to what
extent men and women hold differing preferences
for
particular job
attributes
remains the subject
of
debate,
with a
sizable number
of
empirical studies producing conflicting
results.
These
conflicts may have temporal sources—historical changes
in
men's and women's preferences
for
particular
job
attributes,
as
well
as changes
in
preferences
that
commonly occur over
individuals'
life
cycle.
Most
previous
research has neglected
the
effects
of time
on gender
differences.
Using
data from
national surveys
of
workers over
a
22-year
period,
this
study focuses explicitly on changes by age over
time in
men's and women's preferences for
five
key attributes
of
jobs—short
hours,
high
income,
meaningful
work,
chances
for
promotion,
and
job
security.
The results suggest
that
gender differences
in
preferences have been
both
stable and
limited,
although there is
some evidence
that the
gender gap
in
preferences has actually widened among younger workers
in
recent
years.
Whether men
and
women differ significantly in their preferences for various aspects of work
has
been
the subject of frequent debate by organizational researchers.
The
resolution of this
debate,
in
turn,
has
important
implications for current theoretical disagreements concerning the sources of job
and
occupational segregation by
gender
and,
relatedly, the determinants of earnings differences. It is now widely acknowledged that differences in
the kinds of jobs and occupations held by men and women account for
a
sizable component of commonly
observed differences in average wages
(England,
1992;
Tomaskovic-Devey,
1993).
However, the degree to which
job
and
occupational segregation results from variations in job queues (employees' preferences for different kinds
of jobs) or labor queues (employers' preferences for particular types of workers for given jobs)
is
still
an
unsettled
question
(Reskin
&
Roos,
1990).
And it is
an
important
question.
If women prefer certain types of
jobs,
then their disadvantage in the
marketplace—in terms of relative wages
and
status—is
a
consequence,
not of discrimination or structural
constraints, but of their own "choices" (Mortimer &
Kumka,
1982;
O'Connell
&
Betz,
1996;
Reskin
& Padavic, 1994).
Prominent economic theories of inequality, such
as
human
capital,
are predicated on the assumption of gender-
based differences in job
queues.
Filer
(1985),
for
example,
showed that including such job characteristics
as
perceived social
importance,
opportunities for
promotion,
and
time flexibility, among
others,
in models of wage
estimation could account for
a
sizable portion of the gender
gap
in wages among salaried
workers.
He
concluded
that men
and
women do indeed hold different preferences for jobs
based
on
such
characteristics,
and
that
preference-driven job choices are
an
important source of earnings differences.
His research
did not, however,
provide direct evidence of gender differences in
preferences,
and
the evidence on this point in other studies to
date
is
inconclusive for
a
number of reasons.
One particularly notable problem with research on the relationship of gender to job preferences is
a
general
neglect of the effects of time—in terms of both
age and
period.
It is possible that conflicting empirical evidence on
the question of whether men
and
women differ in their preferences stems from researchers' failure to take
temporal variations into account.
Using
data from national surveys of workers over
a
22-year
period,
this study
investigates changes over time in men's
and
women's preferences for five key attributes of
jobs:
short
hours,
high
income,
meaningful work, chances for
promotion,
and
security. We focus specifically on married men and women
because,
in
theory,
differences in job preferences stem in part from the typical division of household
labor;
therefore,
gender differences are expected to
be
most pronounced among this group
(Menaghan,
1991;
Moen,
1992). Our findings provide some support for claims that gender differences in the job preferences of full-time
workers have been limited for many
years,
but our
research
contradicts assertions that existing differences have
narrowed progressively over time. In fact, in recent
years,
and particularly among younger
workers,
the differences
appear to have widened.
PRIOR
STUDIES OF GENDER DIFFERENCES
IN
JOB
PREFERENCES
Early studies of gendered job preferences provided confirmation for conventional beliefs about why men and
women seek paid employment. Workers' responses to attitudinal
surveys,
conducted
as
part of the well known
Hawthorne studies in the
1930s,
for
example,
indicated that men were more likely to express concern with pay
and working conditions
as
key aspects of their work, whereas women emphasized social relations with colleagues
(Roethlisberger &
Dickson,
1939/1975, p. 245). The researchers interpreted these differences
as
reflecting men's
typical role
as
primary earner in the family and women's role
as
secondary earner. Many of the studies conducted
from the 1960s through the mid-1970s that focused on men's and women's work values also provided support for
this view (Bartol & Manhardt,
1979;
Brenner & Tomkiewicz,
1979;
Centers &
Bugental,
1966;
Manhardt, 1972).
In
a
study of employees in
a
large manufacturing company in the early
1970s,
for
example,
Schuler (1975)
reported that, after taking
age,
education,
and
organizational level into account, women indicated
a
stronger
preference for having pleasant coworkers than did
men,
whereas men attached more importance to having
opportunities to increase earnings
and
to influence important organizational
decisions.
More recent data,
collected by Betz
and
O'Connell
(1989) in
a
study of pharmacy students in the
mid-1980s,
also suggest that men
are more concerned with
income,
security,
and
advancement in
jobs,
whereas women place greater emphasis on
having opportunities to
use
special skills
and
to work with
people.
These
findings are consistent with those from an
analysis of general life values among high school seniors conducted by Beutel
and
Marini
(1995).
Their research
indicated that females were consistently more likely to subscribe to statements that reflect compassion for others
and
a
philosophical commitment to finding meaning in life than
males,
who were more likely to endorse
statements reflecting the value of material benefits
and
competition.
By the
mid-1970s,
however, in the context of growing suspicions about the validity of traditional
presumptions about the nature
and
sources of gender differences (Maccoby &
Jacklin,
1974),
an
increasing
number of studies
began
to question
such
research findings
and
to provide evidence that contradicted
them.
Thus,
for
example,
Spitze
and
Waite (1980) argued that preferences for types of work
are,
to
a
large
degree,
shaped
by
the kinds of jobs individuals hold and by their early work
experiences.
In this light, gender differences in work
values could be
assumed
to result from differences in the occupations
and
positions typically held by women and
men.
Consistent with this argument,
Brief,
Rose,
and
Aldag's (1977) analysis of survey data from
a
national sample
of full-time employees in 1974 showed that once occupation was taken into account, the effects of gender on
preferences for job characteristics were non-significant. Similarly,
a
study by Gomez-Mejia (1990) of employees in
a single corporation indicated that the effects of occupational differences on work attitudes
and
preferences were
much more pronounced than gender differences (see also Brief &
Aldag,
1975;
Brief &
Oliver,
1976;
Lacy,
Bokemeier, &
Shepard,
1983;
Saleh
&
Lalljee,
1969).
Along the same
lines,
Bielby and Bielby
(1989),
using
data
from
a
1977 national survey of workers, found that
men
and women engaged in similar work
had an
equal level of
commitment to work.
A
number of studies have suggested that at least some of the contradictory findings from research on the
relationship between gender
and
work attitudes may be
a
reflection of recent shifts in women's attitudes toward
work, which
have
accompanied large-scale societal changes in gender roles
(Erez,
Borochov,
&
Mannheim,
1989;
Morgan &
Carney,
1985;
see also
Moen,
1992).
For
example,
Brenner
and
Tomkiewicz (1979)
used
data collected
from
a
sample of business school graduates in the late 1970s to replicate research originally conducted by
Manhardt (1972) between 1966
and
1970 on gender differences in job
orientation.
Based
on the comparison of
findings reported in the earlier study and their own
analyses,
Brenner
and
Tomkiewicz (1979) cautiously concluded
that gender differences in evaluations of the importance of various job characteristics were slowly
disappearing.
In
a similar
vein,
Florentine
(1988),
using data on attitudes of college freshmen obtained from surveys conducted
annually between 1969
and
1984,
drew much stronger conclusions about the convergence of men's
and
women's
work values:
The evidence
of
an increasing congruence
in the
values and career expectations
of the
sexes is
significant
if
only
to
suggest
the
extent
that the
female gender role has changed since
1970.
Should
this
trend continue
at
such
a
vigorous
pace,
by
the
century's end
it
may
not
be
an
exaggeration
to
describe
the
degree
of
change
in the
female sex role as something close
to a
"cultural
revolution."
(p.
157)
The empirical evidence of
such
purported shifts in job values over time provided by various studies is
problematic, however. Most of the relevant research
has
been
based
on samples of college-age
and
younger
respondents (Beutel &
Marini,
1995;
Brenner & Tomkiewicz,
1979;
Florentine,
1988;
Lueptow,
1992).
Such
analyses cannot reveal whether observed changes are limited to younger, more recent members of the labor force
or reflect shifts that have affected all
age
groups over time
(i.e.,
are due to cohort or period effects;
see
Ryder,
1965). Moreover, previous studies have shown that the work attitudes
and
values of younger workers are more
apt to shift over time than are those of older workers
(Finn,
1986;
Gomez-Mejia,
1990;
Posner &
Powell,
1990;
Spitze &
Waite,
1980).
Whether this is due to the fact that younger workers are more likely to experience
significant changes in then work environments (because job
and
occupational shifts are more common at younger
ages),
or that older workers react
less
to changes in their work environment is not clear
(Lorence,
1987;
Lorence &
Mortimer,
1985).
In any
case,
evidence of relative instability in younger workers' attitudes
and
values over time
raises the question of whether patterns of work values found among younger respondents hold up at other stages
of the life course.
At the same
time,
work claiming to demonstrate systematic changes in work values across different age
groups
is
also problematic.
Such
work typically
has
been
based
on cross-sectional data (Hackman &
Oldham,
1976;
Hall &
Mansfield,
1975;
Holley,
Feild,
& Holley,
1978;
Loscocco,
1989;
Parker & Chusmir,
1990;
Porter,
1963);
with
such
data,
it is impossible to determine whether observed
age
differences are really due to
aging
processes per se
or, in
fact,
reflect cohort differences.
Even
when longitudinal data are
available,
the potentially differing effects of
cohort membership
and
biological
age
are often ignored
(e.g.,
Jurgensen,
1978).
Even
more sophisticated analyses of gender differences in job preferences, which control for age
and
time
period effects simultaneously (Rowe &
Snizek,
1995),
have yet to explore the extent to which gender differences
might vary both with
age
and over time. Direct examination of time-related changes in gender-based preferences
for job characteristics is
necessary,
however, to address the question of whether differences in
such
preferences
have narrowed only in recent years and/or only among particular
age
groups,
or whether the existence of
differences in men's
and
women's job preferences
has
in fact
been
mythical for many years.
JOB VALUES AND
GENDER:
THE EFFECTS OF AGE AND
TIME
We investigate this question in the following
analyses,
using data from national surveys of adult workers
conducted over
a
22-year
period.
In these
surveys,
respondents were
asked
to give their ranking of five job
attributes— high
income,
job security, meaningful content, opportunities for advancement, and having short
hours—as important aspects of
jobs.
High
income,
job security,
and
opportunities for advancement are all
attributes that involve extrinsic
rewards,
which some evidence suggests are more highly valued by men;
meaningful content, on the other
hand,
is associated with intrinsic rewards, which may be valued more by women
(Lueptow,
1996;
see also Beutel &
Marini,
1995).
The
traditional division of labor within the family is also likely to
lead women to value short hours more than
men,
because women usually feel the stress of integrating work and
family life more than men do
(Menaghan,
1991;
Moen,
1992).
Gender differences in preferences may vary over the life
course,
however.
A
number of studies have
suggested that job preferences shift
as
individuals
age.
For
example,
Wright
and
Hamilton's (1978) analysis of the
1972-1973 Quality of Employment Survey data indicated that younger workers were significantly more likely to
attach importance to opportunities for promotion
as a
job characteristic than were older workers, whereas the
latter were more likely to set
a
premium on job security
and
limited working
hours.
Studies by
Aldag and
Brief
(1975)
and
Cherrington,
Condie,
and England
(1979),
which indicate
a
significant positive effect of
age
on the
degree to which work is viewed
as
having moral importance, suggest that preferences for jobs with meaningful
content may increase with
age.
As
noted previously, however, many studies indicating
age
effects have
methodological limitations that
make
interpreting
such
effects difficult.
Lorence (1987) suggests that there are at least three potential explanations of
an
observed relationship
between
age and
work attitudes. First, it is possible that
such a
relationship reflects cohort differences.
Explanations premised on the concept of cohort assume that job market conditions
and
common work
arrangements that exist when individuals begin their work careers have
an
enduring impact on their work
orientations. Major changes in these conditions
and
arrangements result in distinctive
cohorts,
because newer
labor market entrants presumably develop work orientations that are consonant with the changed conditions,
whereas those who entered the labor market in
an
earlier era will hold orientations reflective of the earlier
conditions. Apparent differences among
age
groups in
attitudes,
then,
may actually be the result of such variations
among cohorts.
Although the notion of cohort effects is intuitively
appealing,
with few exceptions
(e.g.,
Elder,
1974),
most
empirical studies supporting the idea that work attitudes vary significantly by cohort have defined cohorts on
methodological grounds
(i.e.,
based
on regularly
spaced
time intervals,
such as
1900-1909, 1910-1919,
etc.) rather
than on theoretical
grounds;
this practice makes it difficult to impute substantive interpretations to findings of
cohort differences
based such
measures.
Moreover, there is very limited evidence that individuals' work attitudes
do in fact crystallize in
a
way that makes attitudinal change unlikely in the face of changing work environments
(see Lorence & Mortimer,
1985).
Therefore, we find cohort explanations of
age
effects to be less plausible than the
other two explanations that Lorence (1987) describes.
A
second potential explanation of observed
age
differences in attitudes suggests that
such
differences result
from typically age-based reward systems that operate in
many,
if not
all,
organizations (Dannefer,
1984;
Lawrence,
1984).
As a
consequence of
age
grading in the assignment of jobs
and
rewards,
jobs typically held by younger
workers have significantly different characteristics
(e.g.,
less
autonomy
and
lower wages) than those held by older
workers, and these characteristics may be crucial in shaping work attitudes
and
values.
In line with this argument,
Lorence's (1987) analysis of data from
a
two-wave national study of workers showed that
a
positive relation
between job involvement
and age
found among male respondents lessened at one time point
and
disappeared
altogether at
a
later time point, once such job rewards
as
autonomy,
prestige,
and
income were
controlled.
A
third
explanation for
age
differences in work attitudes involves what Lorence
(1987,
p. 536) and others have referred to
as ontogenetic developmental aging
processes.
Embedded in this concept
and
explanation is the notion that
common life course-related
roles,
linked
to
age and
involving social relations outside
as
well
as
inside the
workplace, affect individuals' work attitudes and values.
Our proposed explanation of the relationship between
age
and work values incorporates elements of both of
these latter explanations
and
rests on two key assumptions.
The
first is that age-graded reward
and
promotion
systems in organizations shape individuals' expectations of the kinds of job rewards they are likely to receive
(Dannefer,
1984;
Lawrence,
1984);
the second is that individuals are apt to attach greater importance to job
rewards that they feel they are most likely to
receive.1
Thus,
for
example,
younger workers are likely to evaluate
having opportunities for promotion more highly than older workers, because organizations do in fact promote
younger workers more frequently than older workers
(Lawrence,
1984;
Rosenbaum,
1984;
Sofer,
1970).
Similarly,
younger workers' prospects for substantially improving their earnings are generally much higher than those of
older
workers;
thus,
younger workers are likely to place
a
higher value on the earnings power of jobs.
With few
exceptions,
existing research is largely silent on the issue of whether age-based shifts are equally
likely to occur among men
and
women.
Lorence's (1987) analysis of the effects of
age
on job involvement
indicated that men's job involvement increased up to
a
certain
age,
then
declined,
but for
women,
job involvement
only increased with
age.
Some
indirect evidence bearing
on
the question of gender differences in work attitudes by
age is also provided in
a
study by
Glenn,
Taylor,
and Weaver
(1977),
based
on data from the General Social Survey
for
1972-1974.
Their research suggested that men's level of job satisfaction was more likely to vary with age than
women's.
But to our
knowledge,
no work
has assessed
how gender differences in preferences for job attributes might
change with
age.
Given
our arguments about the influence of career expectations on job preferences, in
combination with extensive evidence on gender differences in career patterns (in terms of earnings
and
mobility),
it seems plausible to expect that men's preferences will increasingly resemble women's preferences
as
they age.
That
is,
women's preferences for jobs that provide
a sense
of personal accomplishment may reflect the
expectation that their jobs are unlikely to provide them with opportunities for substantially increasing earnings or
for promotion. Men's career expectations of holding jobs with
such
attributes are likely to decline with
age,
leading
them to hold preferences that are increasingly similar to those of
women.
Generally, shifting career expectations
with
age
could be expected to
lead
older respondents of both sexes to give highest rank to meaningful content and
security
as
preferred job attributes.
In
addition,
as
noted,
a
number of studies have suggested that whatever
gap
may have once existed in men's
and women's job preferences
has
narrowed considerably over time (Brenner & Tomkiewicz,
1979;
Erez,
Borochov,
&
Mannheim,
1989;
Florentine,
1988).
Although some contrary evidence exists on this point (Beutel & Marini,
1995),
it seems reasonable to investigate the possibility that historical changes in the United States during the time
period of our
study,
from the mid-1970s to the mid-1990s, have led to
an
increasing convergence in gender-
differentiated job preferences over time.
A
number of studies have documented
a
general shift in gender role
ideology, evident from the 1960s
and
early 1970s
onward,
that supported women's increased participation in the
labor force (Mason &
Bumpass,
1975;
Mason,
Czajka,
&
Arber,
1976;
Moen,
Erickson,
& Dempster-McClain, 1997;
Spitze &
Huber,
1980;
Thorton,
Alwin,
&
Camburn,
1983).
This ideological shift,
as
well
as
pressures created by the
depressed economy in the
1980s,
contributed to increases in women's overall level of labor force participation and
to their growing rates of entry into nontraditional jobs
and
occupations throughout the
1970s, 1980s,
and early
1990s
(Moen,
1992).
It seems reasonable to expect that the kinds of changes in gender role ideology
and
women's
labor force participation that occurred over this 20-year
span
would affect patterns of job preferences, resulting in
an increasing convergence in the preferences of men
and
women.2
In this context, we explore the following hypotheses:
1.
Men are more likely to
assign
top rank to the job characteristics of high
income,
job security,
and promotion opportunities than
women,
whereas women are more likely to
assign
top rank
to meaningful content
and
short hours.
2.
Gender differences in job attribute preferences will narrow with
age.
As
career trajectories
become more
settled,
men
are likely to place greater value
on
intrinsic rewards
and
less
emphasis on obtaining high income
and
promotions, whereas women are likely to become
increasingly concerned with job security
as
they grow
older.
3. Gender differences in job attribute preferences will also narrow over
time,
across all age
groups,
partly
as a
result of the increasing similarity in men's
and
women's patterns of labor
force participation over the past two
decades,
and partly
as a
result of general changes in
gender role ideologies that have accompanied the changes in patterns of participation.
These hypotheses reflect
an
underlying assumption that gender differences in job attribute preferences are
not due to cohort differences, that
is,
to the growing representation of a new, younger cohort of women in the
labor
force,
whose attitudes toward work differ substantially from those of earlier cohorts of women employees.
Although we cannot dismiss the possibility that,
as
new cohorts of more egalitarian men
and
women move into the
labor
force,
period changes are preceded by cohort
changes,
there is no evidence that we know of to suggest the
existence of significant, enduring cohort differences in job
values.
However, we consider this possibility in our
discussion of the findings.