Notesale: Turn your study into money

Document Preview

Extracts from the notes are below, to see the PDF you'll receive please use the links above

---

Oxford University Press 2005
All rights reserved
#

Oxford Economic Papers 57 (2005), 479–496 479
doi:10
...
Bender*, Susan M
...
Heywoodz
*Department of Economics, University of Wisconsin-Milwaukee and
Centre for European Labour Market Research, University of Aberdeen, UK
yGraduate Program in Human Resources and Labor Relations,
University of Wisconsin-Milwaukee
zDepartment of Economics, University of Wisconsin-Milwaukee,
P
...
Box 413, Milwaukee, WI 53201, USA, and Department of Commerce,
University of Birmingham; e-mail: heywood@uwm
...
Confirming earlier
studies women report higher job satisfaction than men and higher job satisfaction
in workplaces dominated by women workers
...
Thus, women in female
dominated workplaces may report higher job satisfaction because they value job
flexibility and so choose to dominate the workplaces that provide job flexibility
...


1
...
The recent focus on job satisfaction is motivated
by the recognition that jobs entail complicated bundles of characteristics
...
While a subjective notion,
job satisfaction has been argued to provide a more comprehensive measure of
workers’ utility from the job (Clark and Oswald, 1996)
...
The estimated gender earnings
gap indicates women are paid less, but women in both countries appear more

480

job satisfaction and gender segregation

satisfied with their jobs than men1 (Clark and Oswald, 1996; Clark, 1997; Sloane
and Williams, 2000; Sousa-Poza and Sousa-Poza, 2000)
...
First, satisfaction is a function of expectations, and if women
have lower expectations about labour market outcomes, their expectations are
more easily fulfilled (Clark, 1997)
...

These same two explanations have been used in exploring the role of occupational segregation
...
Yet, much of the literature assumes that women
are ‘crowded’ into a particular set of jobs
...
2 Certainly, Richard Anker of the International
Labour Organization summarizes the position of many: ‘Occupational segregation
by sex occurs everywhere, causing labour market rigidity and economic inefficiency, wasting human resources, preventing change, disadvantaging women, and
perpetuating gender inequalities’ (Anker, 1997, p
...
Thus, one might anticipate
that the job satisfaction of women should be lower in the female dominated jobs
into which they are crowded
...
Sloane and Williams (2000) take this
as evidence that women largely ‘choose’ the jobs they dominate arguing that policy
designed to decrease segregation would reduce the job satisfaction of female
workers
...
The result is that women in male dominated jobs are less
satisfied than women not in male dominated jobs
...
It is
that pattern that we wish to confirm (as there has been no US study of it) and its
explanation which we wish to consider in more depth
...
We find strong evidence of such a partial correlation
...

1
Blau and Kahn (1996) provide estimates of the gender earnings gap for a variety of major industrialized
countries
...
5% in Sweden to 37
...

2

Recent evidence for the US suggests that occupational segregation of women is attributable for approximately one-third of the total gender wage differential (Baldwin et al
...


k
...
bender, s
...
donohue, and j
...
heywood 481
more fundamentally, we explore the cause of the relationship between female job
satisfaction and the female share of the workplace, finding that it arises from
a strong gender specific pattern of sorting based on the ‘flexibility’ of the job to
accommodate family issues
...
We suggest
that flexibility is a crucial element in the bundle of job characteristics that appeal to
women and that it plays an important role explaining the relationship between job
satisfaction and gender segregation
...
Section 3 describes the data
and discusses our methodology
...
Section 5 uses unique questions in the data to control for
the flexibility provided by each job
...


2
...
2) makes clear why economists should study the determinants of job satisfaction: ‘Only one measure, the satisfaction that workers derive
from their jobs, might be viewed as reflecting how they react to the entire panoply
of job characteristics
...

It can be viewed as a single metric that allows the worker to compare the current
job to other labour-market opportunities
...
While such self-reported measures of satisfaction have been criticized
as subjective, Blanchflower and Oswald (1999) explain that such measures have
been successfully used for years by social psychologists and that they do correlate
in expected fashions with many objective outcomes
...
, 1998)
...
, 1994)
...
Job satisfaction is higher for the
youngest and oldest workers (Clark et al
...
, 2002), and for the less educated
(Clark and Oswald, 1996)
...

3
See Bryson et al
...


482

job satisfaction and gender segregation

the role of gender as a determinant of job satisfaction
...
Research within occupations
has found women to be more satisfied than their otherwise equal male counterparts
as clergy (McDuff, 2001), scientists (Dhawan, 2000), attorneys (Hull, 1999), and
doctors (Bashaw, 1999)
...
4
The so-called paradox arises because the apparent job satisfaction of women
seems at odds with their lower adjusted earnings, lower chances for employer
provided training (at least in the US—see Lynch, 1992), and lower adjusted probabilities of receiving pensions and health insurance (Heywood, 1989)
...

As a general rule, women place less value on remuneration
...
Moreover, he shows that those workers who identify earnings as the most
important aspect have lower overall job satisfaction
...
Sloane and Williams (2000) for the UK and Donohue and
Heywood (2004) for the US indicate that higher earnings add more to the job
satisfaction of men than to that of women
...
Finally, increased usual hours of work are often
associated with lower job satisfaction of women but not of men
...
(2000) perform a meta-analysis of 31 studies in the field of organizational
behaviour examining the relative preference by gender of job attributes
...
On the other hand, they conclude that women consider good
co-workers, a good supervisor, and the significance of the task to be more
important than men do
...
5 Thus, the extent of
gender segregation might proxy for these underlying job attributes
...


...
(1995)
...
Perlman and Pike (1994) identify such gender socialization as ‘pre-market discrimination’
meaning the resulting preferences would be taken as given even by a non-discriminatory labour market
...


k
...
bender, s
...
donohue, and j
...
heywood 483
As an alternative, the extent of segregation may itself be considered a job attribute
over which workers have preferences
...
Nonetheless, Tsui et al
...
Clark (1997)
presents very similar evidence also for the UK, finding that the gender satisfaction
gap increases with the extent of females in the workplace, a result that flows from
the greater satisfaction of women in more female dominated workplaces
...
Theoretical work in
economics has addressed the issue of the optimal size of a minority within a workgroup
...
Firms themselves may desire a particular gender (or racial) composition in order to appeal to potential customers
...
Social psychologists
have also measured the influence of work group gender composition on the
commitment and on the effectiveness (productivity) of the group
...
More importantly we identify the critical work
and family flexibility variables that drive that relationship
...
We conclude from our evidence that gender composition does not influence
job satisfaction after accounting for flexibility
...
Data and methodology
We specify estimating equations from a representative US sample, the National
Study of the Changing Workforce (NSCW) conducted in 1997
...
The eventual working sample
size from the NSCW is 1854
...

7
In one such study, Knouse and Dansby (1999) conclude that the optimal female composition is
between 11 and 30%
...
The average job satisfaction for men is 3
...
42 for women
...
The underlying latent
variable is assumed to follow a cumulative normal, and the determinants are
estimated by maximum likelihood using an ordered probit procedure (McKelvey
and Zavonia, 1975)
...

The controls largely follow the economics literature on job satisfaction including demographics, employment relations, industrial and regional dummies, and
human capital variables
...

The critical independent variable on gender composition is built from the
relevant survey question
...
The
responses are one of six categories: 0%, 1–25%, 26 to 50%, 51 to 75%, 76 to 99%,
Table 1 Descriptive statistics and definitions: 1997 NSCW
Variable

Definition (mean, std
...
)

Female
Hours
Wage

Equals 1 if respondent is female (0
...
499)
Natural log of usual hours worked by respondent (3
...
304)
Natural log of respondent’s hourly earnings in USD (2
...
738)
Equals 1 if respondent’s education ended with high school graduation
(0
...
455)
Equals 1 if respondent’s education ended with some college but no
degree (0
...
438)
Equals 1 if respondent’s education ended with an associate degree
(usually a two year course) (0
...
271)
Equals 1 if respondent’s education ended with a bachelor degree
(0
...
412)
Equals 1 if respondent completed an advanced degree (0
...
308)
Respondent’s years of tenure with current employer (7
...
01)
Equals 1 if respondent’s establishment employs between 100 and 500
workers (0
...
444)
Equals 1 if respondent’s establishment employs between 500 and
1000 workers (0
...
421)
Equals 1 if respondent’s establishment employs more than 1000
workers (0
...
300)
Respondent’s years of age (39
...
7)
Equals 1 if respondent is currently a union member (0
...
357)
Equals 1 if respondent’s current job is temporary (0
...
213)
Equals 1 if respondent reports having an employer provide pension
(0
...
437)
Equals 1 if respondent reports no health impairments (0
...
364)
Equals 1 if respondent answers positively that she has some control
over starting and finishing times of tasks (0
...
497)
One digit controls
Dummy variables identifying seven regions within the United States

High school
Some college
Associate degree
Bachelor degree
Advanced degree
Tenure
Size100
Size500
Size1000
Age
Union
Temp job
Pension
Health
Time control
Industry
Regions

k
...
bender, s
...
donohue, and j
...
heywood 485
Table 2 Job satisfaction and gender composition in the NSCW

Male
Not at all
1
2
3
Very
4
Sample size

Female
Not at all
1
2
3
Very
4
Sample size

0% female

Between
0–25%
female

Between
25–50%
female

Between
50–75%
female

Between
75–100%
female

Sample
size

0
...
078
0
...
008
0
...
502

0
...
067
0
...
015
0
...
441

0
...
099
0
...
574
129

0
...
413
208

0
...
446
101

390

Between
0–25%
female

Between
25–50%
female

Between
50–75%
female

Between
75–100%
female

100%
female

Sample
size

0
...
101
0
...
029
0
...
439

0
...
043
0
...
010
0
...
436

0
...
061
0
...
427

0
...
473
171

0
...
585
314

471
164

and 100%
...
Thus, the responses
identify for each worker the share of their immediate group that is white female,
for women, and white male, for men
...
There are no women in the 100% male category, and there are no men in the
0% male category
...
Thus, the estimates for each gender involve four dummies identifying the gender composition of the workgroup
...
From the final column of Table 2, women in the NSCW are more likely to
be represented in the most satisfied category 4 and less likely to be represented in

...
As
a first approximation, this structure might be expected to simply bias standard errors toward no result
...

As an illustration, the gender composition variables generated from the NSCW emerge with the expected
partial correlations in earnings equations: the higher the share female the lower the earnings, all else
being equal
...


486

job satisfaction and gender segregation

the less satisfied category 2
...
The
gender differences in categories 1 and 3 are statistically insignificant
...
4
with 3 degrees of freedom rejecting the hypothesis that the distributions are
identical for each gender
...
Results
Table 2 presents the observations arrayed by job satisfaction and gender composition
...
The share of female workers
in the NSCW reporting being ‘very satisfied’ increases monotonically as the share
of males in the workplace declines
...

As an initial examination, we establish that the broad confines of our data
generate results similar to those already shown in the literature
...
Here the explanatory variables include gender, education, tenure, age, union status, marital status,
establishment size, various job characteristics, as well as industrial and regional
dummies
...
Job
satisfaction increases with earnings and decreases with hours
...

Those with health insurance and those with a pension report higher job satisfaction
...

Each of these results confirms general findings in the literature
...

Estimating the results separately by gender reveals substantial differences
...
This difference mimics earlier
studies by indicating the greater importance of earnings in determining the job
satisfaction of men
...
The significant negative coefficients generated
on higher education in the full sample are largely generated from the male

...

P
10
The test statistic is distributed chi-squared and equals n ðoi À ei Þ2 =ei where oi is the observed
i¼1
number of female observations in satisfaction category i and ei is the expected number of observations
in category i if ei the female proportion in that category was identical to the male proportion in that
category (Walpole, 1974, p
...

11

The absence of a highly significant positive coefficient for hourly earnings would not surprise
Groot and Maassen van den Brink (1999) who claim that many empirical studies fail to find that
higher wages lead to more job satisfaction, ceteris paribus
...
a
...
m
...
s
...
1435** (2
...
2681** (2
...
0809* (1
...
2250 (1
...
2780* (1
...
2319 (1
...
3411** (2
...
4192** (2
...
0108 (0
...
0008* (1
...
1431* (1
...
2841** (3
...
2898** (2
...
0015 (0
...
0001 (0
...
1322 (1
...
2539 (1
...
1826** (2
...
2227** (2
...
3111** (5
...
854** (5
...
998** (3
...
4716 (1
...
4**

À0
...
139)
À0
...
604)
À0
...
797)
À0
...
082)
À0
...
974)
À0
...
129)
À0
...
201)
0
...
677)
0
...
067)
À0
...
287)
À0
...
386)
À0
...
414)
À0
...
688)*
0
...
536)
À0
...
580)
0
...
887)
0
...
851)
0
...
634)
0
...
817)
À3
...
410)
À2
...
314)
À1
...
324)
945
63
...
3645* (1
...
2050** (3
...
2947* (1
...
3480* (1
...
2886 (1
...
4727** (2
...
5763** (2
...
0291* (1
...
0013** (2
...
0851 (0
...
3181** (2
...
2354* (1
...
0201 (0
...
0001 (0
...
1812 (1
...
5543** (2
...
1389 (1
...
2608* (1
...
2598** (3
...
695** (3
...
671* (1
...
0786 (0
...
1**

Note: The regressions include both industry and occupation dummies
...
9, p-val ¼ 0
...

*Statistically significant at the 10% level, **Statistically significant at the 5% level
...
The negative association between job satisfaction and firm size is
disproportionately generated from the female subsample
...

We now add the self-reported measures of gender composition that are specific
to the individual’s workplace
...
Table 4 summarizes the results
...
The separate estimations by
gender show a contrasting pattern
...
The two coefficients for the highly female workplaces take
statistically significant coefficients as well
...
Those in the extremes report higher job
satisfaction with the nearly all female workplaces insignificantly different from

488

job satisfaction and gender segregation

Table 4 Job satisfaction estimates with gender composition included
Variables
Female
100% female
75–100% female
50–75% female
25–50% female
0–25% female
0% female
N
Chi-squared

Full sample
0
...
0515
À0
...
2821**
À0
...
3229**

(1
...
303)
(1
...
968)
(2
...
329)

Women

Men

0
...
025)
0
...
689)
0
...
794)
0
...
487)
À0
...
805)

À0
...
476)
À0
...
824)
À0
...
139)

945
70
...
4**

1819
135
...

*Statistically significant at the 10% level, **Statistically significant at the 5% level
...

1
2
3
4
100% female
75–100% female
0–25% female

0
...
015
0
...
043
0
...
082

0
...
405
0
...
568
0
...
432

Note: These projections hold all other controls at their mean levels while altering the gender composition
...
The projections follow the coefficient estimates from the regressions in Table 4
...
Nonetheless, the basic pattern for the men seems to be that
there is higher job satisfaction in the exclusively male workplaces than in any other
and that there are few differences in satisfaction between all the other workplace
compositions
...
, 2001)
...
Table 5 takes all of the controls at
their mean levels with the exception of the gender composition variables
...
These come from the gender specific estimations
reported in Table 4
...

12
Note that because the two extremes, all female and all male, are collinear with the female dummy, that
dummy is limited to take a value of one when the respondent is a female in a mixed gender workplace
...
a
...
m
...
s
...
In the NSCW, the most female category consists of 57% at the highest satisfaction level compared to 43% in the least
female category
...
14 Estimating wage equations by gender, a predicted
wage for each worker was included in the job satisfaction equations as a
‘comparison wage’ (Clark and Oswald, 1996)
...
Neither alteration changed the pattern presented
...
Flexibility, working conditions, and job satisfaction
Deardorff and Stafford (1976) examine a team production environment where
workers give up flexibility and must cooperate in terms of work times, effort,
and conditions
...
Empirical
researchers have taken the view of team production as foregone flexibility showing
that workers engaged in team production earn more, other things equal (Idson,
1995)
...
Thus, the
estimated relationship between job satisfaction and gender composition might
merely reflect women sorting into jobs that provide flexibility
...

We follow the recognition by Blau et al
...
360–1), that flexibility
includes both the formal policies of a firm and the informal ‘culture’ of the firm
as manifested by attitudes toward, and individual accommodations for, worker
needs
...
Four questions directly probe the
use of time and the extent to which jobs provide flexibility in work scheduling to
meet family responsibilities
...

13
Table 5 presents projections only for those gender composition categories with satisfaction levels
significantly different from that of the base category
...
Unfortunately,
the BHPS does not include information on job flexibility, so we cannot test whether the increased job
satisfaction by women in predominantly female jobs is caused by more flexibility in those jobs
...
Also the
occupational dummies were placed directly in the estimations when the comparison wage was excluded
and the pattern on the indicators of gender composition does not change
...
McBride (2001) details past attempts to confirm the role of comparison earnings
and isolates econometric issues involved in using a predicted comparison wage
...


490

job satisfaction and gender segregation

Table 6 Flexibility variables from the NSCW (means by gender)
Flexibility variables
Sup
...
’ Equals one if strongly agree, zero
otherwise—(Means: Female 0
...
641)

Flex
...
’
Equals one if strongly disagree, zero otherwise—(Means: Female 0
...
320)

Off sick

‘Are you allowed to take a few days off to care for a sick child without
losing pay, without using vacation days, AND without having to
make up some other reason for your absence, or not?’—Equals
one if yes, zero otherwise (Means: Female 0
...
242)

Off hard

‘How hard is it for you to take time off during your work day to
take care of personal or family matters?’—Equals one if not hard
at all, zero otherwise—(Means: Female 0
...
367)

Family or job

‘At my place of employment, employees have to choose between
advancing in their jobs or devoting attention to their family or
personal lives
...
432, Male 0
...
Respondents were asked to
agree or disagree with the statement, ‘At my place of employment, employees
must choose between advancing in their job or devoting attention to their family
or personal lives’
...
16 In some cases the responses were given in a Likert scale that we truncate
to dichotomous as explained
...

Table 6 also shows the mean responses to each of the five questions separately
by gender
...
17 The means of the variables are often similar across genders

...
Thus, the more freedom workers have to design their
own work and working schedules, the greater flexibility they have in meeting competing demands of
work and home (Lombard, 2001)
...
In general, we
searched for all questions that measured the extent of flexibility and include all of those we identified
...
In short, men would purchase more
flexibility from their higher average earnings
...
a
...
m
...
s
...
accommodates
Flex
...
1360* (1
...
0585 (0
...
0833 (0
...
1221 (0
...
1966 (1
...
1982 (1
...
5733** (9
...
4275** (6
...
2400** (3
...
2597** (3
...
2**

Women

0
...
1698
0
...
1170

(1
...
153)
(0
...
739)

0
...
423)
0
...
466)
0
...
864)
0
...
958)
945
167
...
1381 (0
...
1379 (0
...
2285 (1
...
1560 (1
...
6031** (6
...
3621** (3
...
2075** (2
...
2207** (2
...
8**

Note: All controls listed in Table 6 are included in these estimates
...


with three of five showing larger means for the women
...

Table 7 includes indicators except the last and most general summary question of
whether the workplace requires a choice between job and family
...
The coefficients are
large indicating that workers place a high value on flexibility
...
However, the coefficients on the gender composition variables
show a profound change
...
The dummy variable for female continues to show weak statistical
significance indicating women are more satisfied
...
None of the
gender composition coefficients retain significance in either of the gender specific
estimations
...
As a major reason women desire flexibility is to fulfil household responsibilities, this can be thought of as a specific
element of flexibility that may be valued differently by men and women
...
It remains the case that none of the
gender composition coefficients retain significance
...
accommodates
Flex
...
1064 (1
...
0438 (0
...
0595 (0
...
1015 (0
...
1630 (1
...
1681 (1
...
5387** (8
...
3389** (4
...
2255** (3
...
2475** (3
...
3641** (5
...
0**

Women

0
...
1708
0
...
1595

(1
...
119)
(1
...
982)

0
...
805)
0
...
804)
0
...
693)
0
...
713)
0
...
847)
945
215
...
1248 (0
...
1337 (0
...
2190 (1
...
1456 (0
...
5932** (6
...
3382** (3
...
2026** (1
...
2185** (2
...
0912 (0
...
4**

Note: All controls listed in Table 6 are included in these estimates
...


remain statistically significant
...

The remaining two columns reproduce the estimates separately by gender
...
Indeed, the value placed by women
on the indicators of flexibility is dramatic
...
18 Moreover, despite the similarity
and high correlations between the measures of flexibility, they all are individually
important with four of the five coefficients larger in the female estimation
...
The coefficients that were previously significant decline by
44 and 24%, respectively, and the very pattern of the coefficients reverses
...
Instead, the results are consistent with women
sorting into jobs with flexibility
...

18
The pseudo r-squared measure is the likelihood ratio index and its value in the original female
estimation in Table 7 was 0
...
147
...


The coefficient on the 100% female dummy drops from 0
...
19 in Table 8 and the
coefficient on the 75 to 100% female dummy drops from 0
...
17
...
a
...
m
...
s
...


6
...
Yet, this argument suggests that
the expectations of women do not adapt to actual experience
...
Moreover, the evidence presented shows that job satisfaction of women
is highest in the traditionally female dominated work places, the very places in
which women as a whole have the most experience and should have the most
accurate expectations
...

Under such an interpretation women may be socialized not to anticipate much
satisfaction from work and are thus surprised by the actual experience
...
In addition,
Clark (1997, p
...
As the
difference between men’s and women’s jobs is eroded so will the gender difference
in expectations and job satisfaction
...

We suggest that much of the satisfaction difference associated with segregation
results from the exclusion of determinants of satisfaction
...
These results suggest that policies which mandate changes in
gender composition or equality in the bundles of job attributes associated with
gender composition (including earnings) could lower the job satisfaction of both
men and women
...

Similarly, providing women with additional earnings but forcing them to choose
between home and work reduces their job satisfaction (as an additional dollar
of earnings adds far less to female job satisfaction)
...
In the US, the ‘Take Care
Network’ is a coalition supporting increased family leave for care giving and
increased opportunities for flexible scheduling and job sharing (see www
...
org)
...
This recommendation was accepted by the government
which believes it will create ‘a transformation in culture of the workplace’
(DTI, 2001, p
...
To the extent that this legislates increased flexibility, our results

494

job satisfaction and gender segregation

suggest it may be more valuable to women and may reduce the extent of segregation created by women seeking more flexible employment
...
These reduced earnings would follow if the
provision of flexible arrangements is costly to firms that did not previously offer
them (see Heywood et al
...
Thus, policies to promote flexibility may have the
consequence of a economic transfer between genders
...
(1997) Theories of occupational segregation by sex: an overview, International
Labour Review, 136, 315–49
...
(2001) About Time: Flexible Working, Work and Parents Taskforce, London
...
, Butler, R
...
(2001) A hierarchical theory of occupational
segregation and wage discrimination, Economic Inquiry, 39, 94–110
...
(1999) Gender earnings and job satisfaction: the case of US physicians, Ph
...

Dissertation, UMI Number 9916558, University of Wisconsin, Milwaukee
...
A
...
J
...

Bielby, W
...
(1984) A woman’s place is with other women: sex segregation
within organizations, in B
...
), Sex Segregation in the Workplace, National Academy
of Science, Washington, 27–55
...
and Oswald, A
...

Blau, F
...
, and Winkler, A
...
Prentice Hall, Upper Saddle River, NJ
...
and Kahn, L
...

Bryson, A
...
, and Lucifora, C
...

Clark, A
...

Clark, A
...
, and Sanfey, P
...

Clark, A
...
(1996) Satisfaction and comparison income, Journal of Public
Economics, 61, 359–81
...
and Stafford, F
...

Dhawan, S
...

Donohue, S
...
(2004) Job satisfaction, comparison income and gender:
evidence from the NLSY, International Journal of Manpower, 25, 211–34
...
a
...
m
...
s
...
(2001) Government Response to Recommendations from the Work and Parents Taskforce,
Department of Trade and Industry, London
...
and Maassen van den Brink, H
...

Hamermesh, D
...

Heywood, J
...

Heywood, J
...
(2002) Payment schemes and gender in Germany, Industrial and
Labor Relations Review, 56, 44–64
...
, Siebert, W
...
, and Wei, X
...

Heywood, J
...
S
...
(2002) Job satisfaction and sorting: the role of
government and union jobs? Industrial and Labor Relations Review, 55, 595–609
...
(1999) The paradox of the contented female lawyer, Law and Society Review, 33,
687–702
...
(2001) Why women earn less than men in self-employment, Journal of Labor
Research, 22, 817–30
...
(1995) Team production effects on earnings, Economic Letters, 49, 197–203
...
and Solon, G
...

Kidd, M
...
(2000) Female concentration and the gender wage differential in
the United Kingdom, Applied Economics Letters, 7, 337–40
...
and Squires, G
...

Knouse, S
...
(1999) Percentage of work-group diversity and work-group
effectiveness, Journal of Psychology, 133, 486–94
...
, Corrigall, E
...
, and Ritchie, J
...

Lombard, K
...
(2001) Female self-employment and demand for flexible, nonstandard
work schedules, Economic Inquiry, 39, 214–37
...
and Bose, C
...

Loscocco, K
...
(1991) The organization context of women’s and men’s pay
satisfaction, Social Science Quarterly, 72, 3–19
...
(1992) Private-sector training and the earnings of young workers, American
Economic Review, 82, 299–312
...
(2001) Relative-income effects on subjective well-being in the cross-section,
Journal of Economic Behavior and Organization, 45, 251–78
...
(2001) The gender paradox in work satisfaction and the protestant clergy,
Sociology of Religion, 62, 1–21
...
and Zavonia, W
...


496

job satisfaction and gender segregation

Metle, M
...

Millward, N
...
(1995) Gender segregation and male/female wage
differences, Centre for Economic Performance Discussion Paper No
...

Mulinge, M
...
(1998) Employee job satisfaction in developing countries: the
case of Kenya, World Development, 26, 2181–99
...
(1992) The relationship between satisfaction, attitudes and performance: an
organizational level analysis, Journal of Applied Psychology, 77, 963–74
...
, Russel, R
...
, Roberton, M
...
(1995) The paradox of the
contented working woman in intercollegiate athletics administration, Research Quarterly in
Exercise and Sport, 66, 73–9
...
and Pike, M
...
Manchester University Press, Manchester
...
and Weiss, A
...

Rogers, J
...
, and Kash, T
...

Sloane, P
...
and Williams, H
...

Sousa-Poza, A
...
(2000) Taking another look at the gender/job satisfaction
paradox, KYKLOS, 53, 135–52
...
, Egan, T
...
(1992) Being different: relational demography and
organizational attachment, Administrative Science Quarterly, 37, 554–79
...
E
...

Wootton, B
...




---