Write a 2 page paper (the title page and reference page makes 4 pages) in which you:In your own words, write 2 full pages of content summarizing the reading assignment (60%)Use at least two (2) quality references – one should be your textbook and the other should be the assigned reading. Note: Wikipedia and other Websites do not quality as academic resources. (20%)Your assignment must be typed, double spaced, using Times New Roman font (size 12), with one-inch margins on all sides; citations and references must follow APA or school-specific format. Check with your professor for any additional instructions. (20%)Below the attachmentMen and Women in Fiduciary Institutions: A Study of Sex Differences in Career
Development
Author(s): Robert Cabral, Marianne A. Ferber and Carole A. Green
Source: The Review of Economics and Statistics, Vol. 63, No. 4 (Nov., 1981), pp. 573-580
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/1935853
Accessed: 20-07-2016 14:47 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to The Review of
Economics and Statistics
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
MEN AND WOMEN IN FIDUCIARY INSTITUTIONS: A STUDY
OF SEX DIFFERENCES IN CAREER DEVELOPMENT
Robert Cabral, Marianne A. Ferber and Carole A. Green*
M UCH research has been done in recent years
A number of investigations have found seg-
examining the differentials in earnings of
mented markets for labor, each marked by dif-
men and women in particular occupations.’
ferences in wages and career progression.’ One
While these studies have contributed consid-
study of male workers specifically noted that on-
erably to our understanding of this earnings gap,
the-job training and type of starting position are
they tend to accept the occupations as given and
crucial because they are the beginning of a dy-
do not examine the extent to which men and
namic process which continually affects the em-
women have a different occupational distribution
ployee’s earnings. The same study showed that
or question the reasons for it.
starting in the high and middle skill sectors is
In some types of jobs this question does not
consistently associated with higher earnings later
arise at the time the person enters the labor mar-
on.3 In this paper we shall focus on the extent to
ket. For example, when highly specialized skills
which men and women are found in different
are required, as is the case for most professions
entry jobs and how this affects their later earn-
and crafts, only those trained in relevant skills
ings.4
are recruited. But there are many other positions
Finding women in entry jobs which do not lead
at the entry level for which only general require-
to high earnings in the long run need not be to
ments, such as intelligence, literacy, motivation,
their disadvantage if their labor force participa-
or, in some cases, physical strength are needed.
tion tends to be intermittent. It has been argued
Theie are also many higher level positions that
(e.g., Johnson and Stafford, 1974; Polachek,
are filled through upgrading and promotion from
1979) that women take jobs that pay relatively
the lower levels. In such labor markets employ-
well initially but do not provide valuable training
ers, and perhaps employees, are likely to have
leading to great increases in earnings because
considerable discretion with respect to place-
they expect to drop out of the labor market in any
ment of particular individuals into either a
case.
ladder-type or a dead-end job. If gender plays a
This hypothesis would be supported by find-
significant role in this decision, any measure of
ings that women are paid more than men with
achievement and rewards that is restricted to
comparable qualifications initially, and hence
initial placement differences within job catego-
earn more for the limited period they expect to
ries may seriously underestimate the effect of
remain in the labor market, even though other
sex-discrimination on subsequent promotions
jobs pay better later on. If, on the other hand,
and earnings.
men earn as much or more to begin with, we
cannot ascribe their higher earnings in later years
Received for publication September 24, 1979. Revision accepted for publication January 14, 1981.
to their greater willingness to invest in their own
training on the job.
* Brown University; University of Illinois, UrbanaChampaign; University of Illinois, Urbana-Champaign, respectively.
Data
I There are a large number of studies dealing with the per-
For a study concerned with the impact of em-
formance and rewards of highly educated professionals in
ployment decisions on the pattern of occupa-
general, and academicians in particular. A representative
sample of these includes the following: Bayer and Astin
(1975); Gordon, Morton and Braden (1974); Ferber and Kor-
dick (1978); Ferber, Loeb and Lowry (1978); Johnson and
Stafford (1974); Reagan and Maynard (1974); Malkiel and
Malkiel (1973). Some work has also been done on other
occupations, such as Buckley (1971); Blau (1977); Rees and
Schultz (1970); and Hamilton (1973). Hamilton specifically
states her purpose: “to isolate pure measures of wage discrimination on the basis of sex, within narrowly defined occupations” (p. 42).
2 Rees and Shultz (1968); Gordon and Thal-Larsen (1969);
Doeringer (1968).
3 Birnbaum (1976).
4 Thus our investigation is relevant to the dual labor market
theory which suggests that “A worker’s first job in the labor
force . . . should predict the sector in which he presently
works with some accuracy,’ and that ” . . . race and sex will
probably serve as fairly accurate predictors of inter-sectoral
allocation as workers enter the labor market’ Gordon (1972,
p. 50).
[ 573 1
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
574 THE REVIEW OF ECONOMICS AND STATISTICS
tional distribution and salaries, the most suitable
data are those on individual establishments.5
They should ideally include the entire work history of every employee. A set of data became
available that had been supplied to a government
agency by three large fiduciary institutions. They
(or seeking out) particular jobs because of their
(ir)regular hours, seasonal nature, (un)pleasant
work environment, demand for physical strength
or exposure to danger. Should the data show that
men and women with similar qualifications are
found in different entry positions which lead to
were voluntarily provided under an agreement
unequal rewards we would need to look for an
that assured the identity of the institutions would
explanation primarily on the demand rather than
be kept confidential.6 The data cover (1) level of
on the supply side.10
education at time of hire,7 (2) prior relevant ex-
perience,8 (3) number of years with present em-
Methods and Analysis
ployer, (4) initial and present occupational cate-
gory,9 (5) initial and present salary, (6) race, and
(7) sex.
Fiduciary institutions are particularly useful
for this research since several of the variables
that are usually thought to influence the hiring,
placement, and promotion of workers are of no
importance here. The first of these is that the
work is nonseasonal, especially when part-time
workers are excluded, and all employees
officially work the same number of hours per
week. Second, the physical work environment is
more or less the same for everyone, at least to
Simple models explaining earnings based on
the human capital approach abound in the literature. For workers in a single occupation they
tend to include education and relevant experience as explanatory variables. When workers in
various occupations are included in the analysis,
a dummy variable for occupational category is
generally used. Sex is also added, and its coefficient is regarded as an approximate measure of discrimination, with the appropriate
caveats to the effect that all possible differences
in performance may not be represented by the
the extent that everyone works indoors in the
data.
same building. Third, none of the work involves
We begin with such a model, and include occupational categories both at the time of entry
and at present as dummy variables. Other variables included are dummy variables for education denoting the highest degree earned (this was
preferable to years of schooling, since the relationship does not seem to be linear), years of
relevant experience prior to current employ-
physical stress or risk. Thus we need not be
concerned about people of either sex avoiding
i Blau (1977, p. 2).
6 One of the shortcomings of such data is that there is no
information on former employees who have left. To the extent that these differ from present employees, and do so
differently for men than for women, there may be some bias
in the use of this sample.
7Education in different fields is not equally valuable for
particular positions, and it is a well known fact that in the past
women were less likely to major in business related subjects.
A number of studies (e.g., Gordon and Strober, 1975;
Robertson, 1978; Steele and Ward, 1974) found, however,
that even women MBAs have not been rewarded to the same
ment, years worked for present employer, and
sex.1″ The relevant regressions are shown in
table 1.12,13
‘? It may be argued that professional and managerial jobs
involve greater pressures and work beyond the officially designated hours. But such positions are also likely to bring
extent as their male counterparts.
additional rewards, such as greater prestige and authority,
8 Defined as such by the present employer. Since the data
were provided to a government agency which was known to
plan an investigation of potential sex discrimination, it is
reasonable to assume that “relevant” was narrowly construed. Clearly. a person who was a file clerk in another firm
would have accumulated ‘relevant” experience for becoming a file clerk with the current employer, but not for becom-
elegant and more private work space, and last, but not least,
expense accounts. Only if it is assumed that women dislike
additional burdens far more, and value the additional rewards
less than men do, can one conclude that supply side factors
play an important role.
II An alternative approach to a single regression with sex
as a variable would be to run separate regressions for men
ing a typist or an accountant.
and women. This would have the advantage of showing the
9 Present occupational category was provided by the institutions. While initial occupational categories were not furnished, both initial and current job titles were provided, so
that it was possible to develop this information. Detailed
occupational categories were not comparable across institutions. Each tended to use its own classification system. Thus,
in order to make comparisons between them, it was neces-
extent to which the reward structure differs for men and
sary to use gross categories.
women. The emphasis in this study is, however, on the extent
to which differences in occupational structure contribute to
the earnings gap, and this is accomplished better by the
method used here.
12 Data for individual institutions are shown separately,
since as mentioned above, they are probably the most useful
for determining the impact of employment decisions on the
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
SEX DIFFERENCES IN CAREER DEVELOPMENT 575
TABLE 1.-MULTIPLE REGRESSIONS WITH LN CURRENT ANNUAL SALARY AS DEPENDENT
VARIABLE INCLUDING INITIAL AND PRESENT JOB
CATEGORIES AS INDEPENDENT VARIABLES
(unstandardized coefficients)
Institution A Institution B Institution C
Level of Educationc
College Degree .18763a .07618a .10016a
Graduate Degree .07363a .03158b .06972a
Previous Experience .00832a .00641a .00284b
Years Worked .02740a .03442a .02449a
Occupational GroupSd
Manager
Ie
Professional
.13921a
.09063a
.12396a
I .11772a .20928a .23541a
Tech.-Sales
I
.09679a
.09649a
.13162a
Manager
.43733a
.41023a
.49830a
Professional
.09002a
Tech.-Sales
.04479a
.21848a
.19459a
.19399a
.03765
Female
-.
16411
a
-.20968a
-.14442a
Constant
8.87860a
9.14835a
9.069540a
R2
..7895
.7905
.7618
N
1,254
1,098
1,174
Significant at the I’, level.
Significant at the 51t level.
cThe omitted variable is “less than a college degree.”
The omitted category is “all other” consisting almost entirely of clerical workers. Service and blue collar workers are included here because there are too few of
them to make separate analysis useful.
I Initial.
The coefficients for education and years of
work are all significant and of a magnitude consistent with other research on earnings (Corco-
When the percentage differences are translated
into dollar terms for the average employee,
women earn $1,392 less in Institution A, $2,282
ran, 1978; Mincer and Polachek, 1974; and San-
in B, and $1,678 in C than men with comparable
dell and Shapiro, 1978).4 The coefficient for
being female is negative and significant for each
of the three institutions. It is assumed to show
the earnings differences for men and women,
qualifications.
holding constant occupation and career paths.
gender also influences placement into entry job
So far our approach has followed the common
practice of taking the initial and present occupa-
tional category of each employee as given. If
and/or subsequent promotion to other positions,
any measure which fails to control for this must
pattern of occupational distribution and salaries. We shall see
later that there are interesting differences between the institutions.
13 It is interesting to note, that when identical regressions
were run for whites and non-whites separately (the number of
the latter were 222, 125, and 280 for the three institutions),
the coefficients for sex were virtually the same for whites as
shown in this table, -0.16440 for Institution A, -0.18195 for
be expected to underestimate the effect of being
female. Table 2 shows the extent to which differences in occupational distribution contribute to
the earnings gap. The coefficients for being female in the regressions where initial and present
job categories are omitted are uniformly greater
B and -0.13420 for C, but were somewhat smaller for non-
than those seen in table 1. In dollar terms the
whites, -0.05073, -0.13205, and -0.11075, respectively.
negative value of being female increases from
This would appear to confirm the widely held view that black
women are relatively less disadvantaged compared to black
men, than are white women compared to white men. The
other coefficients for whites only are also very similar to
those shown in this table, but there are some differences
for non-whites. Specifically, for Institutions A and C, the
coefficients for college degree were considerably larger,
0.30611 and 0.32362, respectively, but those for most of the
occupational categories were smaller and/or not significant.
14 Information on marital status was not available in this data
set. Fortunately, there is evidence in previous studies that
marital status has no significant effect after work history has
been adequately taken into account. This was found for clerical workers (Ferber and Birnbaum, forthcoming) and for
faculty (Ferber and Green, 1981).
$1,392 to $2,969 in Institution A, from $2,282 to
$4,885 in B, and from $1,678 to $3,857 in C.
It is interesting to note that the coefficient for
being female is considerably larger in Institution
B than in A and C, both when occupational cate-
gory is taken into account and when it is left out.
As we shall see later there are also differences
between the individual institutions in the extent
to which women are at a disadvantage both at the
time they are hired and later on when promotions
have had an effect.
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
576 THE REVIEW OF ECONOMICS AND STATISTICS
TABLE 2.-MULTIPLE REGRESSIONS WITH LN CURRENT ANNUAL SALARY AS
DEPENDENT VARIABLE, EXCLUDING INITIAL AND PRESENT JOB
CATEGORIES AS INDEPENDENT VARIABLES
Institution A Institution B Institution C
College Degree .41297a .25121 a .33872a
Graduate Degree .16153a .10304a .16484a
Previous Experience .01369a .00832a .00586a
Years Worked .03758a .04129a .03683a
Female
-.39025a
.51941a
.37400a
Constant
9.05452a
9.43941a
9.32215a
R2
.6374
.6205
N
1,254
1,098
.5004
1,174
I Significant at the 1′, level.
The fact that men and women are found in
paid, followed by professionals, technical and
different occupational categories when first
sale workers, and last “other,” consisting of
hired, and at present, is to some extent explained
mostly clerical and some service and blue collar
by the higher level of education and longer years
workers. Three separate discriminant analyses
of experience of men. The question remains,
were done as shown in tables 3-5. In the first, the
however, whether sex has a significant influence
dependent variable was initial job group. The
after all other variables have been taken into
discriminating variables were whether or not the
account. To answer this we examine the deter-
individual had a college degree, whether or not
minants of membership in each of the occupa-
the individual had a graduate degree, years of
tional groups. Multiple discriminant analysis
relevant previous experience, and sex. For each
provides a tool for analyzing such a model with a
of the three institutions, at least two significant
polytomous variable.
discriminant functions were constructed. Results
Discriminant analysis constructs discriminant
are shown in table 3. Since the coefficient for sex
functions up to a maximum of the number of
is relatively large in each case, sex is a significant
separate groups minus one, each of which repre-
variable in initial job placement, even when the
sents linear combinations of the criterion vari-
other variables have been taken into account.
ables. The coefficient of the discriminant func-
The related classifications are shown in the
tion is constructed so as to maximize the sepa-
second part of table 3. Only 81%, 79% and 65%
ration between groups. The standardized
of the observations were correctly classified for
coefficients of the discriminant functions show
Institution A, B and C, respectively. This does
the relative weight of each variable in differ-
not appear very impressive as compared with a
entiating between groups. Using the discriminant
naive prediction that all employees are in the
functions, a classification function for each group
dominant “clerical, service, and blue collar”
can be derived. The observations are grouped
category, which would place 79% of cases cor-
according to the classification function. Actual
rectly for A and B, and 61% for C. But this does
and predicted group membership can then be
not give us the full picture. Because virtually all
compared. Examining the number of observa-
women in A and B, 97% and 95% respectively,
tions for which the predicted group coincides
and 79% in C, initially are employed in this job
with the actual group, relative to what would
group, the naive forecast inevitably does ex-
occur by chance, provides a measure of the suc-
tremely well for them. The results are quite dif-
cess of the discriminating variables in predicting
ferent for men where the comparable figures are
52%, 51% and 40%.
group membership.
In this case we specify four values for the
The second and third discriminant analyses
dependent variable, each representing member-
were run for present job placement. For both, the
ship in one of the occupational groups: 1 = man-
additional discriminating variable, seniority,
ager; 2 = professional; 3 = technical and sales; 4
was added. For the third, initial job category was
= other. There are distinct differences among the
three institutions in mean earnings betveen the
also used. Again, sex is shown to be a highly
significant variable, particularly when present
four occupational groups. Managers are highest
job placement is examined with initial job cate-
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
SEX DIFFERENCES IN CAREER DEVELOPMENT 577
TABLE 3.-INITIAL JOB GROUP
Standardized Discriminant Function Coefficients
Institution
A
Institution
B
Institution
Cb
Variable Function I Function 2 Fuinction 3 f’-value FLunction I Function 2 F-value Function I Function 2 f’-value
Pr-evious
ex
i-ience
–
0.23
-0.0(9
0.90
15.61
-0.t
-0.80
9.75
0.2()
College deg -ee -).76 0.70 -1).02 160.86 -0.61 0.01 58.85 –0.78 0.35 104.11
Graduate degr-ee –(0.32 0.0045 0.29 21.27 -0.38 0.57 25.94 -0.34 –0.87 18.90
Sex” -11.54 —-(.79 -11.29 77.74 -0.66 -0.16 74.68 -0.66 -0.1)4 69.77
Percenta ge ot variance
expl tined by the
function 96.15 2.72 1.13 94.72 4.89 98.24 1.27
Level of significance 0.()0 0.0() 0.00 0.00 0.00 0.00 0.02
Classification Matrices
Institution
A
Institution
Actual
B
Institution
Actual
C
Actual
Predicted Group Member-ship Number Predicted Gr-oup Membership Number Predicted Gr-oup Membership NumberCler-icalt
Service,
Clerical,
Service.
Clerical.
Ser-vice,
Tech/ Blue Tech/ Blue Tech/ Blue
Actual Group Manager Pl-of. Sales Collar Manager- Pr-of. Sales Collar Manager Prof. Sales Collar-
Manager 22 42 2 29 95 5 13 0 7 25 0 74 10
Professional I ) 68 3 26 107 6 72 6 48 132 0 94 15
70 154
135 244
Technical
0
and
sales
5
11
2
48
66
7
19
3
42
71
0
27
38
65
Clerical. service.
blue
collar
15
23
19
929
186
7
58
22
783
870
0
43
2
666
711
Totals 52 144 26 1,032 1,254 25 162 31 881) 1.098 0 238 27 909 1.174
Percentage of
observations
corriectly
classified
81.42
78.60
64.74
Note: Only discriminant functions significant at the 11.05 level or better tre r-eported.
‘ Sex her-e is a dummy variable: I = matle: 0 = female.
b Pr-evious experience proved to be too insignificant to be included in the discriminant functions for this institution.
TABLE 4.-CURRENT JOB WITHOUT INITIAL JOB GROUP
Standar-dized Discriminaint Function Coefficients
Instituition
A
Institution
Bb
Institution
C
Function Function Function Function Function Function Function Function Function
Vatriable
Previous
1
2
3
F-value
experience
-0.12
1
-0.39
2
3
-0.08
F–value
5.69
0.42
1
2
0.10
3
F-value
0.22
0.04
2.72
College degr-ee -0.61 -0.35 -0.74 81.34 -0.45 0.32 -0.48 37.29 -0.65 -0.63 0.07 73.41
Gr-aduate degr-ee -0.26 -0.11 0.15 13.38 -0.16 0.37 -0.74 7.01 -0.27 -0.28 0.72 12.45
Sexa
-0.65
0.60
0.43
105.71
-0.81
-0.04
0.50
154.70
-0.61
0.28
-0.60
56.23
Year-s of seniority -0.43 -0.73 0.48 48.32 – 0.38 -0.82 -0.47 33.04 -0.42 0.63 0.51 37.00
Percentage of variance
explained by the
function
93.16
5.14
1.70
96.19
3.13
0.67
84.68
13.98
1.34
Level of significance 0.1)0 (1.00 0.00 0.00 0.00 0.04 0.00 0.00 0.03
Classification Matrices
Institution
A
Institution
Actual
B
Institution
Actual
C
Actual
Predicted Group Membership Number Predicted Group Membership Number Predicted Group Membership Number
Clerical,
Ser-vice,
Tech/
Blue
Clerical,
Service,
Tech/
Blue
Clerical,
Service.
Tech/
Blue
Actual Group Manager Prof. Sales Collar Manager Prof. Sales Collar Manager Prof. Sales Collar
Manager 198 () 0 145 343 180 13 0 54 247 97 49 0 129 275
Professional 29 0 0 33 62 94 15 0 20 129 24 94 0 168 286
Technical and sales 6 0 0 35 41 53 5 0 17 75 3 19 0 42 64
Clerical. service,
blue
Totals
collar
272
0
41
0
0
980
0
767
1,254
808
411
84
44
11
0
0
643
552
647
1,098
24
148
18
180
Per-centage of
observations
correctly
classified
76.95
68.03
59.45
Sex here is a dummy variable: I = male: 0 = female.
Previous experience proved to be too insignificant to be included in the discriminant functions for this institution.
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
0
0
507
846
549
1,174
578 THE REVIEW OF ECONOMICS AND STATISTICS
TABLE 5.-CURRENT JOB WITH INITIAL JOB GROUP
Standardized Discriminant Function Coefficients
Institution
A
Institution
B
Institution
C
Function Function Function Function Function Function Function Function Function
Variable
Previous
College
degree
Graduate
Sex
1
2
-0.23
degree
-0.44
3
F’-value
experience
-0.09
0.01
-0.30
-0.12
-0.16
-0.08
-0.02
1
-0.15
20.88
-0.16
44.21
2
0.28
5.09
0.63
3
-0.11
0.01
0.05
-0.06
F’-value
3.90
0.42
–0.36
0.06
0.01
1
-0.10
14.61
-0.34
85.96
3
0.26
2.21
0.21
2
F-value
-0.08
-0.13
0.08
0.09
0.15
0.01
–0.09
2.79
12.18
0.15
-0.64
2.18
21.51
Initially manager -0.50 -0.05 -0.30 48.38 0.19 0.12 0.54 10.96 0.68 –0.03 0.28 74.28
Initially professional -0.48 -0.10 0.76 61.39 0.46 0.34 -0.09 40.20 0.83 -0.01 0.43 127.95
Initially
tech/sales
-0.45
0.90
0.01
235.66
0.24
-0.87
0.07
71.44
0.24
0.95
-0.12
91.74
Years of seniority -0.38 -0.23 -0.48 52.07 0.43 0.19 0.63 44.65 0.26 -0.21 -0.61 30.13
Percentage of variance
explained by the
function
Level
of
63.72
30.66
significance
5.62
0.00
79.24
0.00
17.80
0.00
0.00
2.96
0.00
74.05
0.00
19.30
0.00
6.65
0.00
0.00
Classification Matrices
Institution
A
Institution
Actual
B
Institution
Actual
C
Actual
Predicted Group Membership Number Predicted Group Membership Number Pr-edicted Group Membership Number
Clerical,
Service,
Tech/
Blue
Clerical,
Service,
Tech/
Blue
Clerical,
Ser-vice,
Tech/
Blue
Actual Group Manager Pr-of. Sales Collar Manager P-of. Sales Collar Manager Prof. Sales Collar
Manager 138
Professional
Technical
58 25 122
8 28 0 26
and
sales
I
343 94 44 21 88 247 184 28 8 55
62 42 40 4 43 129 89 72 20 105
1
32
7
41
7
3
35
30
75
4
4
30
275
286
26
64
Clerical, service,
blue
Totals
collar
167
87
20
65
0
8
935
780
1,254
808
170
27
87
0
70
10
771
610
647
1,098
298
21
5
109
8
66
515
701
549
1,174
Percentage of
obser-vations
correctly
classified
77.99
70.95
68.23
Sex here is a dummy variable: I = male. 0 = female.
Previous experience proved to be too insignificant to be included in the discriminant functions for this institution.
gory omitted. The resulting standardized discrim-
this hypothesis is to run a multiple regression
inant functions are shown in tables 4 and 5.
with initial earnings as the dependent variable,
Classification matrices for current job without
and level of education, years of job experience
and with initial job group are also reported in
and sex as the independent variables. A sig-
tables 4 and 5. It can be seen that discriminant
nificant positive coefficient for being female
analysis does far better than the naive prediction
would tend to support the hypothesis. A sig-
that all employees are in the “clerical, service,
nificant negative coefficient, on the other hand,
and blue collar” category. Correct placement
would point toward discrimination against wom-
constitutes 77%, 68% and 59% for the analysis
en in the form of lower earnings from the
without initial job category and as high as 78%,
beginning of their work-life onward. Table 6 (a
71% and 68% with initial job category, as com-
and b) shows that the latter is the case, whether
pared to only 65%, 55% and 47% for the naive
or not initial job category is entered as an inde-
prediction.
pendent variable.
In other words, women and men with compa-
Since previous relevant experience is used as a
rable qualifications are placed in different entry
variable, it may, however, be argued that men
positions, and those in similar entry positions are
are placed in different job categories because
differentially promoted afterwards. As a result
they have accumulated their experience in differ-
there are significant differences in present job
ent fields. It is also entirely possible that women
category between men and women who were
have accumulated their experience over a longer
initially hired with comparable qualifications.
We turn next to the question whether men earn
period of time, because they are more likely to
interrupt their labor force participation. Our data
more in later years at least in part because they
do not enable us to test these possibilities di-
accept entry positions that have greater upward
rectly. Therefore we examined a subsample con-
mobility but pay less initially. A simple test of
sisting only of persons with no previous experi-
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
SEX DIFFERENCES IN CAREER DEVELOPMENT 579
TABLE 6.-MULTIPLE REGRESSIONS WITH LN INITIAL SALARY AS DEPENDENT VARIABLE
Institution
A
Institution
B
Institution
C
(a)
Including Initial Job Category as Independent Variable
College
Degree
.26669a
.10587b
.12367a
Graduate Degree .12307a .04156b .11087a
Previous Experience .21423a .14886a .0414lb
Manager
I
.25870a
.21505a
.36017a
Professional
I
.25157a
.41310a
.46545a
Tech.-Sales
Female
–
.
Constant
R2
N
I
.20272a
15192a
–
8.66884a
.5563
1,254
.19798a
.23321a
8.92882a
.5129
1,098
.16026a
–
14140a
8.94908a
.4643
1,174
(b)
Excluding Initial Job Category as Independent Variable
College Degree .42641 a .26582b .33699a
Graduate Degree .1 8179a .13116a .20242a
Previous
Female
Experience
.25801a
–
–
.28480a
.17817a
.41338a
–
.03218
.32123a
Constant
8.77991
a
9.09544a
9.16197a
R2
.4777
.3758
.2836
N
1,254
1,098
1,174
(c)
Excluding Initial Job Category as Independent Variable, for
Those With No Previous Experience
College
Degree
.51731a
.26436a
.32553a
Graduate Degree .16318a .15241a .18540a
Female
.19803a
.40697a
.32607a
Constant
8.67571a
9.06856a
9.16895a
R2
.4021
.3443
.2747
N
684
701
996
Significant at the 1’S, level.
b Significant at the 574, level.
ence. As table 6(c) shows, the negative co-
ous institutions. It is therefore interesting to note
efficient for being female remains significant in
that the size of the coefficients is quite different
all three institutions. Furthermore, it does not
between the various regressions for the three
even decrease greatly in two out of the three
institutions in tables 1, 2, and 6. The fact that the
cases. Thus, differences in previous experience
coefficient for being female is negative in all
between men and women do not adequately ex-
cases does suggest that some of the unexplained
plain lower earnings of women in their initial
differential may be caused by real differences
between men and women, but the substantial
jobs.
One objection that is invariably raised when
differences in the size of the coefficient, as well
regressions are used to determine the contribu-
as the variation in earnings differentials at time of
tion that a variety of measurable differences in
hire and in promotion policies, point toward the
qualifications make for the explanation of the
likelihood that institutions practice varying de-
earnings gap between men and women is that
grees of several types of sex discrimination.
there are likely to be other differences also which
cannot, or at any rate have not, been measured.
The regressions used in this study are as open to
Summary and Conclusions
this criticism as any others. To the extent, how-
Our investigation shows that male employees
ever, that these unmeasured attributes of men
in the three fiduciary institutions have more edu-
and women exist, they should cause the co-
cation and experience, but that the differential in
efficient for sex to be about the same for vari-
earnings is considerably greater than can be ac-
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
580 THE REVIEW OF ECONOMICS AND STATISTICS
counted for by these differences. Men are paid
more than women with comparable qualifications
in the same occupational categories. Men are
also more likely to be placed into higher job
categories than women with a comparable educa-
Doeringer, Peter, “Manpower Programs for Ghetto Labor
Markets,” Proceedings, Industrial Relations Research
Association, 21st Annual Meetings, 1968.
Ferber, Marianne A., and Bonnie J. Birnbaum, “Labor Force
Participation Patterns and Earnings of Women Clerical
Workers,” forthcoming, Joiur nial of Humnan Resources.
tion, and their pay in these jobs is higher from the
Ferber, Marianne A., and Carole A. Green, ‘Equity in Hir-
beginning. These differences are found to exist
ing: A Case Study of a Large Public University,”
even though virtually all the jobs in our sample
are indoor, white collar jobs which require no
special physical strength or impose physical
hardships, and, with the exception of some professional jobs, require no advanced education or
specialized training in previous jobs. This is clear
because in all categories people without these
qualifications are frequently hired.
The extent to which women are paid less than
men varies considerably between institutions, as
does the extent to which they are placed in various job categories, either when first hired or later
on. While it is rather clear that in terms of earn-
ings women fare worse relative to men in Institution B, the most important finding is that the
extent and type of discrimination differ in each
institution. For if women were treated differently
from men because of real differences between
them, albeit differences not measured by the
variables, one would expect the patterns to be
similar between essentially similar institutions
with similar job structures.
Thus our data provide evidence that women
experience differences both in terms of pay and
job assignments and that these differences are at
least to some extent arbitrary. The most plausible conclusion is that they are caused by sex
discrimination.
University of Illinois at Urbana-Champaign, College of
Business Administration, Faculty Working Papers
#747 (1981).
Ferber, Marianne A., and Betty Kordick, “Differentials in
the Earnings of Ph.D.’s,” Industr-ial and Labor Relationis Review 31 (Jan. 1978), 227-238.
Ferber, Marianne A., Jane W. Loeb, and Helen M. Lowry,
‘The Economic Status of Women Faculty: A Reappraisal,” Jouiii-ai of’ Huinan Resources 13 (Summer
1978), 385-401.
Gordon, David MI., 7’iheor-ies of’Poviertv (itd Underenmplov-
ment (Lexington, MA: Lexington Books, D. C. Heath
and Co., 1972).
Gordon, Francine E., and Myra H. Strober, Bringing Womeni
into Management (New York: McGraw Hill Book
Company, 1975).
Gordon, Margaret S., and Margaret Thal-Larsen, Em plover
Policies in a Changin,g Labor- Market (Berkeley: Institute of Industrial Relations at the University of California, 1969).
Gordon, Nancy M., Thomas E. Morton, and Ira C. Braden,
“Faculty Salaries: Is There Discrimination by Sex,
Race or Discipline?” American Economic Review 64
(June 1974), 419-427.
Hamilton, Mary, ” Sex and Income Inequality among the
Employed,” The Annals of the Americaln Academy
of Political aind Social Science (Sept. 1973), 42-52.
Johnson, George E., and Frank P. Stafford, “The Earnings
and Promotion of Women Faculty,” American Economnic Review 64 (Dec. 1974), 888-903.
Malkiel, Burton, G., and Judith A. Malkiel, “Male-Female
Pay Differentials in Professional Employment,” American Economic Review 63 (Sept. 1973), 693-705.
Mincer, Jacob, and Solomon Polachek, “Family Investments
in Human Capital: Earnings of Women,” JournaIl of
Political Economy 82 (Mar/Apr 1974), S76-108.
Polachek, Solomon, “Occupational Segregation: Theory,
Evidence and Programs,” in Cynthia B. Lloyd, Emily
S. Andrews and Curtis L. Gilroy, Women in the Labor
Market (New York: Columbia University Press,
1979).
REFERENCES
Bayer, Alan E., and Helen S. Astin, “Sex Differentials in the
Academic Reward System,” Science (May 1975),
796-802.
Birnbaum, Howard, “Career Origins, On-the-Job Training,
and Earnings,” Soulthern Economic Journal 42 (Apr.
1976), 587-599.
Blau, Francine, Equal Pay in the Office (Lexington MA:
Lexington Books, D. C. Heath and Co., 1977).
Buckley, J. E., “Pay Differences between Men and Women
in the Same Job,” Monthly Labor Review 94 (Nov.
1971), 36-39.
Corcoran, Mary, “The Structure of Female Wages,” American Economic Review 68 (May 1978), 165-170.
Reagan, Barbara, and B. Maynard, “Sex Discrimination in
Universities: An Approach Through Internal Labor
Market Analysis,” Anmer-ican Association o( University Pro fessors Bulletin (Spring 1974), 11-21.
Rees, Albert, and George P. Shultz, Workers and Wages in
an Urban Labor Market (Chicago: University of
Chicago Press, 1970).
Robertson, W., “Women M.B.A.’s, Harvard ’73,” Fortune
(Aug. 28, 1978), 50-60.
Sandell, Steven H., and David Shapiro, “The Theory of
Human Capital and the Earnings of Women. A Reexamination of the Evidence,” Journal of Hunman Resources 13 (Winter 1978), 103-117.
Steele, J., and L. Ward, “M.B.A.’s: Mobile, Well Situated,
Well Paid,” Harvard Business Review (1974), 99-1 10.
This content downloaded from 132.174.254.243 on Wed, 20 Jul 2016 14:47:37 UTC
All use subject to http://about.jstor.org/terms
Purchase answer to see full
attachment
Why Choose Us
- 100% non-plagiarized Papers
- 24/7 /365 Service Available
- Affordable Prices
- Any Paper, Urgency, and Subject
- Will complete your papers in 6 hours
- On-time Delivery
- Money-back and Privacy guarantees
- Unlimited Amendments upon request
- Satisfaction guarantee
How it Works
- Click on the “Place Order” tab at the top menu or “Order Now” icon at the bottom and a new page will appear with an order form to be filled.
- Fill in your paper’s requirements in the "PAPER DETAILS" section.
- Fill in your paper’s academic level, deadline, and the required number of pages from the drop-down menus.
- Click “CREATE ACCOUNT & SIGN IN” to enter your registration details and get an account with us for record-keeping and then, click on “PROCEED TO CHECKOUT” at the bottom of the page.
- From there, the payment sections will show, follow the guided payment process and your order will be available for our writing team to work on it.