Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 27
A Woman's Place Is Win Other
Women Sex Segregation
Within Organizations
WILLIAM T. BELBY and
JAMES N. BARON
Sex segregation in the workplace is one
of the most visible signs of social inequality.
In almost every work setting, it is unusual
to see men and women working at the same
job. When they do, they typically perform
different tasks, with unequal levels of re-
sponsibility and authority. Even when job
tasks are virtually identical, it is not uncom-
mon to find men and women allocated to
distinct job classifications within an orga-
. .
n~zat~on.
Even women working full time, year round
are paid less than men. While the earnings
gap is partly due to unequal access to high
wage firms and to unequal pay for compa-
rable work, a substantial portion is due to
differences in pay scales for job cIassifica-
tions filled by men and those Filed by women
(Bridges and Berk, 1974; Halaby, 1979;
Treiman and Hartmann, 1981~. Sex segre-
gation has social-psychological conse-
quences as well. For example, groups with
limited opportunities for advancement may
respond with psychological disengagement
from the firm, Towered career aspirations,
and an increasingly narrow, instrumental
orientation toward work (Kanter, 1977b).
In short, sex-segregated workplaces affect
27
us personally. Social structures that gener-
ate gender segregation are of great concern
to social scientists, and the inequities that
segregation engenders are obviously rele-
vant to social policy. Yet sociologists know
surprisingly little about job segregation by
sex. Most of what we have learned concerns
segregation among occupations. For exam-
ple, we know that equalizing the detailed
(census 3-digit) occupational distribution for
men versus women would require moving
roughly 60 percent of women working out-
side the home across occupational cate-
gories, and this has changed very little since
1900 (Gross, 1968; Blau and Hendricks, 1978;
Williams, 1979; England, 1981a). We are
also learning more about how men and
women make occupational choices (Bielby,
1978; Marini, 1978, 1980~. Empirical re-
search on job segregation across organiza-
tional settings, however, is quite sparse.
Accordingly, this paper examines sex seg-
regation in the workplace, utilizing data de-
scribing work arrangements in nearly 400
establishments across a wide range of in-
dustrial and institutional settings. We dis-
tinguish situations in which employers place
men and women in the same job cIassifi-
OCR for page 28
28
WILLIAM T. BIELBY AND JAMES N. BARON
cation from those in which job titles seg-
regate the sexes within establishments.
Our study could be viewed as a straight-
forward job-level disaggregation of findings
regarding occupational sex segregation. Those
studies acknowledge that considerable seg-
regation may exist even within detailed oc-
cupational categories. Our measures and
methods parallel such studies, and our find-
ings confirm their speculations about per-
vasive segregation within occupations.
However, our aim is not merely to reveal
hidden segregation among jobs within firms.
Rather, since sex segregation is accom-
plished in organizations and is affected by
technical, administrative, and social exigen-
cies of the workplace, it is important to ex-
amine how organizational structures and
processes produce sex segregation.
Our research does not consider how men's
and women's occupational choices, labor force
participation, and human capital invest-
ments affect the sex composition of the
workplace (for contrasting interpretations,
see England, 1982, and Polachek, 19791. Nor
are we investigating the demand side in the
economist's sense of the term, since we have
no information on the productivities of dif-
ferent classes of workers and the wages em-
ployers are willing to offer them (cf. Blau,
19771. Rather, the intersection of labor sup-
ply and demand enters into our analysis in-
directly, since occupational composition and
skill mix of the firm are examined as deter-
minants of sex segregation. However, if jobs
in most establishments are highly segre-
gated by sex—even across firms differing
dramatically in their production functions
and cost structures then it seems unlikely
that marginal adjustments of supply and de-
mand account for distinct job assignments
of men and women (for a similar view, see
Blau and Jusenius, 19761.
WHY ARE SOME FIRMS MORE
SEGREGATED THAN OTHERS?
Diverse explanations of sex segregation
have been reviewed thoroughly by others
(e.g., Blau and Jusenius, 1976; England,
1981b). Much more has been written, how-
ever, about why employers treat men and
women differently than about the extent to
which they do so. The sparse literature ad-
dressing why some firms are more segre-
gated than others falls into three categories:
institutional accounts, explanations based on
tastes for discrimination, and human capital
market models.
Institutional accounts stress how statisti-
cal discrimination in hiring and allocating
employees places men and women in dis-
tinct career trajectories. Men tend to enter
internal labor markets in which they can
expect an orderly progression through suc-
cessively more attractive jobs, insulated from
competition outside the firm. This increases
organizational loyalty, decreases costly
worker turnover, and allows employers to
recoup investments in firm-specific training
(Doeringer and Piore, 19711. Women are
perceived to have weaker commitments both
to specific firms and to paid employment in
general and are thus allocated to jobs with
low turnover costs and limited opportunity
for security and advancement (Bielby and
Baron, 19821. Not all firms, however, re-
quire specifically trained workers or have
internal labor markets. Therefore, if sex biases
in allocating workers to job ladders were the
only basis for segregating men and women,
one would expect less segregation in firms
lacking institutionalized employment ar-
rangements—particularly small, labor-in-
tensive, entrepreneurial firms in the so-called
economic periphery (Averitt, 19681.
This is certainly not the only mechanism
placing men and women in distinct job cIas-
sifications, and perhaps a more reasonable
hypothesis is that the process of segregation
differs according to an organization's admin-
istrative arrangements and location within
the economy. For example, small manufac-
turing, service, and retail establishments
typically rely on an unskilled secondary la-
bor market and use simple hierarchy or en-
trepreneurial despotism to control workers
(Edwards, 19791.
OCR for page 29
SEX SEGREGATION WITHIN ORGANIZATIONS
29
These enterprises might apply one of sev-
eral strategies in allocating men and women
to jobs. Highly trained line workers with
job- and firm-specific skills typically are not
employed in such establishments, nor are
highly rationalized personnel and job cIas-
sification procedures utilized. Thus, these
firms might provide precisely the work con-
texts in which men and women who lack
credentials for more desirable employment
work together within broadly defined job
categories. Furthermore, if employers must
sacrifice profits in order to discriminate, they
must be able to afford the costs of their pol-
icies. Marginal firms with weak competitive
positions can least afford these costs and have
an economic incentive to ignore sex in hiring
and allocating workers (Arrow, 19731.
In the absence, however, of institution-
alized procedures for hiring and allocating
workers, male employers in the economic
periphery may have more discretion to im-
plement tastes for discrimination, which can
reflect their own preferences or those of their
employees or even their clients. In the most
extreme case, patriarchal control strategies
would exclude women from the workplace
entirely. Such arrangements should be most
prevalent in organizational niches that are
protected from competitive pressures (e.g.,
through satellite linkages with larger firms)
or where preferences for a segregated work
force are so widely held within an industry
or area that they have the force of customary
law constraining market forces (Doeringer
and Piore, 1971:22-271.
If labor supply and technical requisites de-
ter~ine the distribution of men and women
across job categories, then a firm's mix of oc-
cupations and skids should largely account for
its tendencies to segregate men from women.
According to such human capital models,
workers expecting intermittent labor force
participation (primarily women) choose to en-
ter occupations in which job skills do not atro-
phy from nonuse (Polachek, 19791. Indeed, if
jobs with the highest turnover costs are also
those in which skills atrophy most quickly,
then extreme segregation can reflect maxi-
mizing behavior by both workers and em-
ployers. That is, firms wiD assign men and
women to the same job titles only under spe-
cific, and rare, circumstances: (a) when there
is an available labor pool composed of men
and women and (b) when employers perceive
that the costs of employing men and women
roughly are the same.
To summarize, certain analysts argue that
gender segregation at work is caused by ad-
ministrative arrangements for hiring, allocat-
ing, and controlling employees. Others em-
phasize the impact oftastes or prejudices, while
still others claim that sex segregation reflects
rational decisions regarding human capital in-
vestments on the part of workers and em-
ployers. Perhaps because segregation is such
a natural attribute of most work situations,
little has been written about the conditions
under which it does not occur.
Our empirical analysis is guided by sev-
eral general hypotheses. First, institutional
accounts suggest that less segregated firms
lack the administrative apparatus to differ-
entiate workers by sex and cannot afford the
costs of implementing employers' tastes for
a segregated work force. Second, neocIass-
ical accounts, grounded in notions of tech-
nical efficiency, suggest that desegregated
organizations do not rely heavily on firm-
specifc skills but employ workers in occu-
pations that are attractive to both men and
women and for which both sexes are eligi-
ble. Of course, each of the mechanisms sum-
marized above might operate but within
specific organizational settings. Conse-
quently, we examine the heterogeneity
among highly segregated establishments to
see if there are alternate strategies by which
employers achieve the same result: distinct
job assignments for men and women.
DATA AND METHODS
We analyzed data on work arrangements
in hundreds of economic establishments stud-
fed in California between 1959 and 1979 by
the California Occupational Analysis Field
Center ofthe U.S. Employment Service. These
OCR for page 30
30
WILLIAM T. BIELBY AND [AMES N. BARON
data, used primarily in preparing the D'c-
t~onary of Occupational Titles (DOI), and
procedures for collecting them are described
in U. S. Department of Labor (19721. Our unit
of analysis is the establishment, the "physical
location where business is conducted or where
services or industrial operations are per-
formed"(U.S. Bureau ofthe Census, 1976:iv).
Industrial characteristics serve as contextual
Actors, and suborgan~ational information about
workers and jobs has been aggregated to char-
acterize each enterprise. The majority of the
establishments are firms; others are branches,
regional divisions, subsidiaries, and produc-
tion sites. Since we focused on work sites rather
than firms, corporate headquarters of multi-
plant organizations are typically not included
in our data. Corporations often direct initial
desegregation efforts at headquarter mana-
gerial and office work (Shaeffer and Lynton,
1979), and progress toward equal employ-
ment opportunity (EEO) goals in these areas
will not be reflected in our results.
The Sample
No well-defined sampling frame guides
the Employment Service's selection of en-
terprises to study, but they try to represent
the diversity of activities carried out within
any industry (Miller et al., 1980~. The Cal-
ifornia Field Center tended to study those
industries that are regionally concentrated
in the state, so our sample of establishments
includes, for example, firms engaged in ag-
riculture, aircraft manufacturing, banking,
fishing, and motion picture production but
not automobile or furniture manufacturing.
While our sample provides a reasonable
representation of the composition of estab-
lishments within industries, the actual in-
dustries studied are not fully representative
of economic activities in California. Most
importantly, manufacturing establishments
~ Reweighting our observations according to pub-
lished data on the size distribution of establishments
within California's industries has virtually no effect on
the distribution of organizational attributes in our sam-
ple.
are overrepresented in the sample. Major
California industries not represented in our
sample include construction trades, truck-
ing, apparel and general merchandise retail
trade (department stores), and insurance
carriers. Ike first two industries are male
dominated and highly segregated; the latter
two employ many women and may be less
segregated. While these data do not char-
acterize a distinct population, they do re-
flect a diversity of work arrangements across
a broad range of industrial and organiza-
tional contexts. In our view they provide
invaluable comparative evidence regarding
how administrative, technical, and environ-
mental contingencies in organizations affect
the structuring of work.
The data collected and coded for our proj-
ect include 742 observations in over 500 clis-
tinct enterprises. About one-fifth of the es-
tablishments were visited more than once
by Employment Service analysts. The most
recent analysis was used for firms with fol-
low-up data. Since some of the information
used to characterize organizational attri-
butes, however, was derived from narrative
reports (described below), precedence was
given to complete observations that also
possessed a contemporaneous narrative re-
port, even if a more recent follow-up anal-
ysis, lacking a narrative, had occurred.
To ensure comparability, analyses re-
stricted to the firm's productive component
or some other subset of jobs or departments
were omitted, since they do not accurately
characterize an entire work site. This re-
striction reduces the sample of establish-
ments to 415. The sex composition of jobs
was not reported for 22 of these firms, re-
ducing the sample size for analyses reported
in this paper to 393.2 Of these, about 26
2 Job composition was not enumerated by sex after
1977-1978, apparently because of increasing resistance
from establishments approached by the Employment
Service. Unfortunately, this occurred when the Cali-
fornia Field Center was studying agricultural estab-
lishments; therefore, 7 of the 22 observations lacking
information on sex composition are in agriculture.
OCR for page 31
SEX SEGREGATION WITHIN ORGANIZATIONS
31
percent are plants, franchises, or production
sites within some larger entity (though not
necessarily corporations); 10 percent are ad-
ministrative divisions, regions, or branches
of larger companies; 3 percent are subsidi-
aries (distinct firms owned by another firm);
and 61 percent are independent busi-
nesses.3 Over half, 54 percent of these es-
tablishments were studied between 1968 and
1971, and 76 percent were visited by the
Employment Service between 1965 and 1973.
The 393 establishments in our sample em-
ploy nearly 47,000 men and over 14,000
women.4
The Documents
Lois paper uses two types of data obtained
from the records of the California Occupa-
tional Analysis Field Center. Staffing Sched-
3 In practice, it sometimes was difficult to determine
precisely if establishments studied by the Employment
Service were autonomous firms or productive or ad-
ministrative units within larger companies. When our
materials indicated an owner or president, we assumed
the enterprise was autonomous, owner-operated, un-
less other information indicated to the contrary. When
the top position had such titles as plant superintendent,
plant manager, general manager we assumed the en-
terprise was a subdivision of a larger firm, unless back-
ground information suggested otherwise. Anomalous
cases were referred for clarification to the Employment
Service analysts who conducted the original studies.
Confidentiality restrictions precluded access to estab-
lishments' identities, preventing us from resolving such
ambiguities directly.
4 The disproportionate share of manufacturing estab-
lishments in our sample accounts for the underrepre-
sentation of women workers. Nevertheless, the range
of industries covered represents nearly every work con-
text in which women labor. One important exception:
The Employment Service tends to analyze branch plants
and to overlook corporate headquarters. Therefore, vir-
tually every kind of nonmanual work performed by
women is represented in our study, but, unfortunately,
we have no instances of such work done at the head-
quarter of fices of large corporations. Evidence from the
early 1970s suggests that efforts to desegregate non-
manual work occurred first in such contexts (Shaeffer
and Lynton, 1979~.
ules supply, in essence, a complete organi-
zational division of labor for the plant or firm
in which job titles are analyzed. Face Sheets
provide iden~ing information about the es-
tablishment and analysis. After assigning the
firm to one or more categories of the Standard
Industrial Classification (SIC), the Employ-
ment Service classifies the enterprise by its
primary products and supplies information
about any unique or noteworthy characteris-
tics of the firm, such as its jobs and processes.
Narrative reports prepared for many estab-
lishments include information on some or all
of the foldowing: history and purpose of es-
tablishment; environmental conditions; op-
erations and activities (departmentation,
workBow, processes or services); personnel
policies and practices; recent job restructur-
ing, erects of automation on personnel and
operations; and optional sections dealing with
such topics as the product market and rela-
tions with government, the community, or
other firms.
Operationalization
Staking schedules, face sheets, and nar-
rative reports were used to measure various
environmental, organizational, and techni-
cal attributes of establishments as well as
the composition of occupations and skills
employee! in each enterprise. Operational-
izations are summarized in Table 3-1, which
also reports descriptive statistics for varia-
bles used in our analysis. Organizational scale
is measured by the natural logarithm of the
number of employees, and positional spe-
cialization is measured by the logarithm of
median job size. Ike latter measure is com-
puted across workers, so a median of 10 in-
(licates that one-half of the workers are in
establishment job titles with 10 or more in-
cumbents as opposed to half the jobs con-
taining 10 or more workers. This measure
indexes the degree to which establishments
"massify" the work force by assigning many
workers to the same job title. Consequently,
low scores correspond to high levels of spe-
OCR for page 32
32
WILLIAM T. BIELBY AND JAMES N. BARON
TABLE 3-1
(N = 393)
Descriptive Statistics and Operationalizations of Organizational Attributes
Variable (range)
Establishment Scale and Specialization
I~og establishment size(.69-8.97) 3.67
Log specialization (~6.49) 1.12
Manufacturing Industry (~1) .63
Standard
Mean Deviation
N Description
Economic Sector
Core (0-1)
Ambiguous (0-1)
Periphery (0-1)
Social Organization
Log fragmentation ~—1.10-1.39)
Union or bidding arrangements
(~1)
Proportion women (.00-1.00)
Occupational and Skill Composition
Proportion production workers
(.0C~1.00)
Proportion clerical and sales
workers (.0~1.00)
Proportion service workers
(.00-1.00)
Proportion professional, mana-
gerial, technical workers
(.00-1.00)
1.48 393 Natural log of number of employees in estab-
lishment.
1.03 368 Natural log of median job size (see text).
393 One or more of establishment's standard indus-
trial classification designations is in the man-
ufacturing industry.
.38 393
.41 — 393
.21 — 393
.14
.25
.32
.56
.20
.08
.16
Average complexity: data 1.92
(0-5.71)
Average complexity: people 1.27
(0-8.00)
Average technical skills -0.01
(-2.19~.08)
Skill specificity (.00-.99) .51
Sex Segregation (12.~100) 93.4
See text.
See text.
See text.
.28
.26
.33
.23
.21
.20
1.09
1.11
1.00
.27
13.8
360 Natural log of ratio of establishment job titles
to unique DOT titles.
393 1 = some or all employees unionized or covered
by formal bidding arrangements.
393 Proportion of female employees.
380
379
379
379
336
336
379
173
393
Proportion of workers with DOT codes denoting
production occupations.
Proportion of workers with DOT codes denoting
clerical and sales occupations.
Proportion of workers with DOT codes denoting
service occupations.
Proportion of workers with DOT codes denoting
professional, managerial, or technical occu-
pations.
Mean of ratings indicating complexity of work-
ers' involvement with data.
Mean of ratings indicating complexity of work-
ers' involvement with people.
Mean of standardized ratings of production
workers' technical skill (see text).
Proportion of workers in nonentry-level jobs.
Index of dissimilarity computed across job titles.
100 = all male, all female, or completely seg-
regated.
cialization.5 One way to segregate workers
is to place each worker in a unique job cat-
egory; therefore, we expect greater sex seg-
regation in more specialized establish-
ments. We Ailed to detect any net differences
in relationships between sex segregation
5 This measure correlated - .87 with Gibbs and Pos-
ton's (1975) "M4" index of distributive differentiation,
describing the evenness of the distribution of workers
across positions. Functional (horizontal) differentiation
of departments is reflected in our criteria for assigning
establishments to economic sectors (see below). Struc-
tural differentiation the proliferation of work roles
is almost completely determined by organizational scale;
the correlation between the number of employees and
the number of job titles is .92 when both are measured
in a logarithmic metric. Consequently, while such dif-
ferentiation may mediate the effects of scale, it is un-
likely to affect work arrangements independently.
OCR for page 33
SEX SEGREGATION WITHIN ORGANIZATIONS
33
across 6 industrial sectors: manufacturing,
state, services, social overhead capital, ag- ~
riculture, and trade. We distinguish man-
ufacturing from nonmanufacturing, how-
ever, to assess indirectly whether physical
demands of work account for patterns of sex
segregation.
We argued above that organizations con-
front different incentives to segregate de-
pending on their niche within the economy
and their size, structure, and technology.
These differences in organizational forms and
environments capture distinctive locations
within the economy what some institu-
tionalists and Marxists have called sectors
or segments. In certain respects, these are
the organizational equivalent of"cIasses";
that is, actors presumed to share certain in-
terests and attributes by virtue of their com-
mon market positions. Core firms are typ-
ically large, differentiated, use automated
technologies, produce multiple products, are
unionized, and are linked to larger organi-
zational entities. Their environments are
characterized by interorganizational de-
pendence within key industrial sectors, and
these establishments tend to be dominant
actors in their milieux. In contrast, the eco-
nomic periphery is composed of small, un-
differentiated enterprises, typically operat-
ing in highly competitive markets in
industries other than manufacturing. This
congruence of organizational form and en-
vironment floes not characterize all firms;
the 161 establishments allocated to the am-
biguous sector include those in small-scale
manufacturing and many nonmanufacturing
firms in less vulnerable situations with re-
gar<1 to their environments. The measures
and procedures underlying this sectoral
classification scheme are discussed in detail
by Baron (1982:Chapter IV).
Radical accounts of workplace relations
suggest that three aspects of an establish-
ment's social organization should be asso-
ciated with its level of sex segregation. Frag-
mentation is measured by the logged ratio
of job titles in the establishment to unique
6-dizit DOT titles assigned by the Employ-
ment Service analyst.6 It measures the de-
gree to which the organization differentiates
its work force administratively beyond what
might be expected from a breakdown of de-
tailed occupational functions (Braverman,
1974:70-83; Edwards, 19791. Fragmentation
is one strategy for segregating male and fe-
male workers who perform similar job tasks;
that is, separate names are attached to men's
and women's work. Other facets of the social
organization of workplaces are measured by
a dummy variable that denotes the presence
of unions or formalized bidcling arrange-
ments, covering some or all workers, and
the sex ratio, the percentage of workers who
are female. Unions and formal bidding ar-
rangements are an institutional arrange-
ment that may constrain employers' (and male
employees') ability to indulge tastes for dis-
crimination. Some argue, however, that
unions can exacerbate gender inequalities
(e.g., Milkman, 1980; Baron and Bielby,
19821. Finally, tokenism—a highly unbal-
anced sex ratio can facilitate the segre-
gation of women or men into one or two
separate job titles, while a more balanced
work force may be more difficult to segre-
gate (Kanter, 1977a).
The occupational composition of each en-
terprise was computed from the DOT oc-
cupational cocles corresponding to each job
titIe.7 We measured the distributions of
workers across the following broad cate-
gories: professional, managerial, and tech-
6 In many firms, not every job title was mapped to
a DOT title by the Employment Service analyst. The
fragmentation measure was computed only for firms in
which both 90 percent of the jobs and 90 percent of
the workers were assigned a 6-digit DOT code (see
Baron, 1982: Chapter VI). Six-digit DOT codes do not
correspond to unique titles in the DOT, but more de-
tailed classification is possible only for those jobs as-
signed to occupational categories according to the fourth
edition classification scheme.
7 Establishments in which less than 75 percent of the
workers could be assigned a DOT code were treated
as missing data on this outcome.
OCR for page 34
34
WILLIAM T. BIELBY AND [AMES N. BARON
nical; clerical and sales; service domestic,
business, personal and production, in-
cluding extractive and transportation occu-
pations. Levels of informational and inter-
personal skill were measured by the mean
of ratings indicating the complexity of work-
ers' involvements with data and people, re-
spectively.8
The measures of technical skills used de-
pended on the edition of the DOT used in
the analysis of an establishment. Production
occupations were classified as skilled, semi-
skillecI, or unskilled in the second edition,
and we assigned values of 2, 1, and 0 to these
categories, respectively. Ratings of workers'
involvements with "things" began with the
third edition. Average ratings were com-
puted for each establishment's labor force
using either the measure based on second
edition DOT codes or the one based on sub-
sequent versions of the DOT. Each estab-
lishment's score was then normalized rela-
tive to all others sharing the same version
of the skill measure. That is, each enter-
prise's level of technical complexity is meas-
ured relative to other organizations incor-
porating the same version of the DOT
occupational ciassification.9 Finally, the level
of firm-specific skills was assessed indirectly
by the proportion of workers in "line" de-
8 This information, when available, was coded from
job analyses pertaining to each position in the firm.
Otherwise, it was obtained from 6-digit DOT occu-
pational codes listed on staffing schedules. Third edi-
tion DOT ratings of 7 or 8 for relations with data were
recoded to 6 to conform to the fourth edition rating
scheme. Mean ratings for each establishment were
computed only if: (a) at least half the jobs and workers
in line departments could be characterized on the data
and people dimensions; (b) at least half of the jobs and
workers in other departments could be characterized;
and (c) no more than 10 percent of the establishment
labor force was missing data on these variables. Finally,
scale values were inverted so that large values corre-
spond to high levels of involvement with data and peo-
ple.
9 The mean level of technical skills was computed
from third or fourth edition ratings (from job analyses
or DOT codes), subject to the same restrictions de-
scribed in footnote 7. Third edition codes of 8 were
recoded to 7 to conform to the fourth edition rating
partments who were not in entry-level jobs,
as indicated on the staffing schedule.~°
The index of dissimilarity measures the
percentage of workers of one sex that would
have to be moved to new job classifications
in order to equalize the job distribution of
workers by sex (Duncan and Duncan, 19551.
It equals zero when the percentage distri-
butions of men and women across job cat-
egories are identical, and it equals 100 when
no men and women work in the same job. ii
The Analysis
We first describe the distribution of es-
tablishments by level of sex segregation. Then
we examine the organizational attributes that
distinguish propensities to segregate. The
relative contribution of social, administra-
tive, and technical attributes to patterns of
segregation may suggest mechanisms that
account for those results, but we do not ex-
pect conclusive results from cross-sectional
findings. Accordingly, these analyses are
supplemented in two ways. First, we ex-
amine specific cases for which qualitative
information exists on the hiring and alloca-
scheme. Mean ratings were computed from second edi-
tion codes only if: (a) detailed occupational codes ex-
isted for at least 75 percent of the employees in the
enterprise and (b) at least 25 percent of the workers
were in production jobs. Preference was given to third
edition data for establishments that contained jobs an-
alyzed in terms of both second and third edition DOT
procedures.
is Workers in nonproduction departments were ex-
cluded because the Employment Service did not al-
ways collect information denoting entry-level jobs in
those departments. The measure was not computed for
establishments with less than 15 percent of their labor
force in production-related departments. We also elim-
inated observations in which any department had no
entry-level workers, or in which certain traditionally
entry-level occupations occurred e.g., janitor, re-
ceptionist but none was coded as such. Given these
restrictions, a measure of skill specificity is available
for less than half of our observations.
ii We consider all-male and all-female establish-
ments to be perfectly segregated and assign a value of
100 to the index of dissimilarity. However, in analyses
reported below, these enterprises are considered sep-
arately.
OCR for page 35
SEX SEGREGATION WITHIN ORGANIZATIONS
35
TABLE 3-2 Distribution of Establishments and Workers by Level of Segregation
(N = 393)
(4)
(3) Cumulative (5)
(2) Percentage Percentage Median (8)
(1) Number of of of Establish- (6) (7) Cumulative
Segregation Establish- Establish- Establish- ment Number of Percentage Percentage
Index meets meets meets Size Workers of Workers of Workers
~9
10-19
20-29
30-39
40-49
50-59
60-69
70-79
80-89
90-96
9~99.99
100
All male
All female
TOTAL
o
2
2
13
18
27
35
52
202a
21
393
0.0
0.3
0.5
1.0
0.5
1.8
3.3
4.6
6.9
8.9
13.2
51.4
5.3
2.3
100
0.0
0.3
0.8
1.8
2.3
4.1
7.4
12.0
18.8
27.7
40.9
92.3
97.7
100 5
36
30
94.5
64
25.5
11
28
29.5
57
72
195
25
o
30
189
457
51
186
540
3,746
2,281
7,260
26,587
19,250
322
51
60,950
0.0
0.0
0.3
0.7
0.1
0.3
0.9
6.1
3.7
11.9
43.6
31.6
0.5
0.1
100
0.0
0.0
0.4
1.1
1.2
1.5
2.4
8.5
12.2
24.2
67.8
99.4
99.9
100
a Subsequent tables report 201 completely segregated establishments. After completing our analyses, however,
we discovered 1 establishment for which the sole integrated job was in fact due to a coding error.
tion of women. Second, we examine the
subset of organizations for which longitu-
dinal information is available in order to team
(1) the extent to which patterns of sex seg-
regation change over time and (2) the or-
ganizational circumstances under which
gender segregation increases, decreases, or
remains constant. The concluding section
addresses implications of our findings for
policies aimed at equalizing job experiences
and attainments of men and women.
RESULTS
Descriptive Statistics
Table 3-2 presents a remarkable story: Most
establishments are either completely seg-
regated or nearly so. Less than one-f~fth of
the establishments have segregation indices
lower than 90, and they employ less than
one-eighth of the workers in our sample.
Over one-half of the establishments are
completely segregated, and over three-
quarters of the workers are in organizations
having indices between 96 and 100. Indeed,
only 10 percent of the nearly 61,000 workers
are in establishment job titles that have both
men and women assigned to them. Even
among the 162 establishments having some
men and women in the same job titles, the
mean segregation index is 84.1. In short,
the workplace is substantially more segre-
gate(1 by sex than is suggested by studies
that aggregate work force composition across
establishments and into 3-digit occupational
categories. i2
One way to segregate mate and female
workers is to employ either men or women
exclusively in an establishment. The 21 es-
tablishments without female workers, listed
by establishment size in Table 3-3, are al-
i2 of course, statistics on the distribution of workers
in our sample are not representative of the California
labor force, since establishments, not workers, were
sampled. Nevertheless, these results show that there
are very few work contexts in which men and women
are assigned to the same job titles, and results reported
below suggest that even the least segregated enter-
prises are seldom examples of workplace equity be-
tween the sexes.
OCR for page 36
36
_.
o
sit
o
Cal
-
Ct
-
o
o
-
Ct
U)
V'
._
U:
4-
o
in
·50
¢
7
o
.
Ct
.
be
o
~ .
m
C,7
O
Ct an,, ~
'=t .>
4) ~
O
U'
S .C}
O
.
o.= ~0.
Ct ·
=5 ~
~ O
~ _
o
_l Ed
Cp
=0
o
~8
U'
-
o
C)
Cal
o
:~
.>
So C)
en
o
Cal
o
z
U'
Cal
Ct
of
,¢ S t~0 0 Ct e
S ~ C O C
~ C1 S ~ S 0 ~ S
,,Q ~ C
U) ~ ~ C~ ~ P~
¢ <~¢ <<
Z~ZZZZZZZ~ ZZZ^^ZZ~
Z Z
O ° ° ° ~ ° ~ U. Co ° ~ ~ ~ 0 ~ ~ tS C~ 00 ~ 00
_ Z
O O O O O O O O O O O O O O C~ ~ C~ C~
~ _
O O
O O O O ~ O ~ O ~ O O ~ ~ O ~ ~ — d4
oo O O ~ O ~ ~ ~ O ~ ~ ~ 0 ~ '7 ~ ~ u: ~r
_ _ _ _ _ ~
O O C~ O O O O O O CO O O ~ C~ O O O O O ~ O
. . . . . . . . . . . . . . . . . . . . .
C~ _ _ _ _ C~ _ C~ _ C~ _ C~ C] ~ _ C~ CO CO O 00
_ _
U: O ~ O O O O ~ 00 0 ¢ 00 U: O ~ ~ O _ O O CO
. . . . . . . . . . _ . . . . _ . . . . .
_ _ _ _ _ O O _ ~ O O _ O ~ _ _
CD IO ~ ~ ~ O ~ ~
_ _
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 00 0 _ ~ t- CO
c' V V V V ~ V V V V V V V ~ V V
o ~ ~ ~ ~ ~ ~ ~ ~ ~ V¢ ~ ~ ¢V ~ ~ ~ ~
_
C~
CC CO
S
o0 ~ ~ ~
U) U: U) ~
~ ~ ¢ ~ ~ ~ ¢ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ -¢ ~
.o
e e ~ ~ ~ ~ ~ C ~ ~ =- ~ e =^ °~ ~ ' eO
OCR for page 37
OCR for page 45
SEX SEGREGATION WITHIN ORGANIZATIONS 45
TABLE 3-6 (Continued)
Occupational Composition: Percentage in
Entry Job Titles Median Clerical
No. per DOT Job Size Production Sales PTMC Service Comments
Sixteen Least-Segregated Establishments
o
8
1 1.0
0.9
s
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
1.1
0.5
1.8
1.4
1.0
1.0
1.0
1.0
1.0
1.0
1.3
0.8
1.3
1.4
1.9
1.4
1.4
2.4
1.3
1.4
3.2
1.3
NA
1.6
2.1
100
o
67.9 0
6.8 0
25.5 0
59.1 8
2.2 0
1.8 0
11.4 100
4.1 0
1.2 0
1.0 62
2.0 0
1.0 57
2.0 86
2.0 9
8.0
61.2
18.0
NA
111.5
102.6
15.0
268.3
111.7
157.5
69.2
79
35
79
75
15
84
o
8
NA
74
77
7
76
22
82
78
o
o
94
67
12
17
25
o
5
6
54
13
10
60
6
68
o
92
93
24
78
10
22
70
o
5
o
25
80
18
14
70
0 Picker—14M, 14F.
0 Language teachers—6M, 6F; department head—3M,
2F.
o
o
o
1
o
30
o
o
33
o
3
o
o
16
Eleven Largest Establishments
14
3
5
11
7
8
30
33 25
NA NA
26 1
8 15
8
3
4
18
2
2
34
NA
o
o
Apt. manager—23M, 126F.
Salesmen—13M, 4F.
Escrow manager—4M, 24F; escrow officer—3M, 32F.
Salesman—130M, 80F.
Sales clerk—5M, 1F.
Counselor—3M, 3F.
Worker—11M, 12F.
Salesman—1M, 7F.
Salesperson—1M, 1F.
Instrument maker—3M, 1F.
Psychologist—3M, 2F; therapist—13M, 25F; camp
counselor—4M, 3F.
Assembler—4M, 1F.
Four women in 3 integrated production jobs.
Counselor—llM, llF.
No women placed in production jobs due to nature of
work.
Women mostly in clerical and bindery classifications.
Eleven women in 3 integrated quality control classifica-
tions; none in production.
Women mostly in clerical classifications.
See text.
Company policy to place women in other than routine
clerical positions. Several integrated officer and man-
ager jobs.
Women mostly in clerical classifications.
Women employed only in clerical, stewardess, and
ticket agent classifications.
Women employed without restriction except in jobs re-
quiring lifting 25 pounds or more.
No women assigned to production classifications due to
vigorous requirements of various crabcs.
c PTM = professional, technical, and managerial occupations.
supply of male and female workers who are
ill prepared for most other types of em-
ployment, and employers bear none of the
training costs (l~omas, 19801. Neverthe-
less, lhomas's study suggests that the two
agricultural establishments in our sample may
be atypical. He found that men in lettuce
harvesting are typically assigned to higher
paic} piece-rate jobs, while women are con-
centrated in hourly crews. Unfortunately,
we have no evidence of the generalizability
of his or our findings, since the sex com-
position of jobs was not compiled for the
Employment Service's more recent analyses
of agricultural work.
In sum, no single dimension of desegre-
gation emerges from our analysis of"de-
viant" organizations. Instead, we found 4
qualitatively di~erent but sometimes over-
lapping sets of circumstances that contribute
to a desegregated workforce: (1) nominal de-
segregation of a single job title in a very
OCR for page 46
46
WILLIAM T. BlELBY AND JAMES N. BARON
small firm; (2) spatial segregation across work
sites of mate and female workers assigned
to the same job classification; (3) sex-linked
desegregation of jobs like camp counselor,
which employ men and women, but toward
different ends; and (4) market desegregation
that occurs when a mixed labor pool is avail-
able to employers who perceive that train-
ing and turnover costs are identical for male
and female workers. Moderate desegrega-
tion occurs rarely in the work contexts we
have examined, and, when it does occur, it
is typically of the nominal variety.
Large Establishments and Bureaucratic
Segregation
Almost all large establishments are highly
segregated, and most have written rules
governing the hiring and allocation of work-
ers. Consequently, if bureaucratic control
strategies segregate men from women (Ed-
wards, 1979), this should be most apparent
in the largest establishments. The bottom
pane! of Table 3-6 lists attributes of the 11
establishments in our sample employing more
than 800 workers. Case materials provide
insight into administrative roles and pro-
cedures that support sex segregation.
All but one of these establishments seg-
regate employees almost perfectly by sex,
and narrative information available for 7 of
them suggests that this total segregation is
accomplished largely through bureaucratic
rules and procedures. Possibly the most im-
portant factor, particularly in manufacturing
establishments, is the existence of legal re-
strictions on the weight that women may lift.
California law specified until 1970 that "no
female employee should be requested or
permitted to lift any object weighing 50
pounds or over," and regulations enforced
by the state's Industrial Welfare Commis-
sion further restricted the maximum to 25
pounds. In 1970 a federal court ruled that
this law conflicted with Title VII of the Civil
Rights Act (Utility Workers' Union of Amer-
a vs. Southern California Edwon t69-5434~.
References to these restrictions occur re-
peatedly in both narrative reports and job
analyses. For example, the Naval shipyard
in Table 3-6 assigns no women to production
jobs due to "vigorous requirements of var-
ious crafts" (Narrative report 3, 1965), and
the mining enterprise followed the same
policy due to the "nature of work" (Narrative
report 1800, 1971~. A company providing
telephone service to a large metropolitan
area states that it was company policy to
employ women "without restriction" except
in jobs requiring lifting 25 pounds or more,
yet the segregation index was 99.2 (Narra-
tive report 100, 1965~.
The ordnance plant, studied in 1970, em-
ployed 555 females as assemblers and 243
males as production workers. Each is an en-
try job, and both were mapped to the same
detailed DOT occupational category by the
Employment Service analysts According
to the job analysis, these jobs differ pri-
marily in that male "workers" lift 25 to 40
pounds, but female "assemblers" lift 5 to 20
pounds.
Weight restrictions are not mentioned ex-
plicitly in the narrative for the printing es-
tablishment in Table 3-6 (one of the largest
in the western U. S. in 1968), but as in nearly
all other manufacturing plants, the only pro-
duction activities assigned to women are light
assembly tasks. Most production tasks in this
establishment are done by skilled crafts-
men. The union contract establishes pro-
cedures for hiring and apprenticeship, and
it seems reasonable to conclude that the union
plays a substantial role in enforcing sex seg-
regation in this plant. A narrative report
prepared in 1966 for one of the two race
tracks provides evidence of that role in an-
other organizational context: Union dis-
patching policies explicitly exclude women
from the job of parimutuel clerk (Narrative
report 15361. No report was prepared for
the other track in the same area, but pre-
i~ In other words, at the occupational level, entry-
level production work has a relatively balanced sex ratio
in this firm.
OCR for page 47
SEX SEGREGATION WITHIN ORGANIZATIONS
47
sumably it falls under the jurisdiction of the
same union. Of course, these industries might
have been sex-segregated long before they
were unionized. That is, unions may be per-
petuating gender-based inequalities rather
than creating them.
Lifting restrictions and union contracts
cannot account for all sex segregation in large
establishments, because segregation is
equally pervasive in large, nonunion estab-
lishments outside manufacturing. Indeed,
in most firms including all but 1 of the
11 in the bottom pane! of Table 3-6 non-
production jobs are as segregates! as pro-
duction jobs, if not more so. While this sit-
uation primarily reflects distinctions between
male managers and female clerical workers,
it is also true that the few male clerical work-
ers and female managers in our sample of
establishments are hardly ever assigned to
job classifications with workers of the op-
posite sex.
One notable exception is the bank studied
in 1968, which employs more women than
any other establishment in our sample. The
bank claimed it had recently initiated a pro-
gram to hire and promote women into of-
ficer classifications (Narrative report 415,
19681. The staffing schedule supports this
claim: In 1968, women comprised 7 of 81
vice presidents, 18 of 108 assistant vice pres-
idents, 16 of 49 management trainees, and
29 of 118 operations officers. While females
were used exclusively in routine data pro-
cessing jobs like keypunch operator, 10 of
23 systems analysts were women.
In one sense the bank's efforts are only
noteworthy when contrasted! against the
uniformly high levels of segregation in other
comparable large establishments: Fully
equalizing the job distribution by sex would
still require reclassifying 80 percent of this
bank's female employees. Most managerial
and professional positions remained exclu-
sively male, while few men were employed
in routine clerical duties. An organization's
demography, history, technology, and labor
supply, however, constrain the degree to
which its work force composition can change
in a short period of time. This is especially
true of large bureaucracies employing many
workers in nonentry jobs. As long as sen-
iority ant! accrued skills remain important
bases for advancement in such contexts,
workplace equity cannot happen overnight:
The existing stratification regime favoring
males essentially guarantees more work-
place inequality in the short run, just as
reduction of childbearing to below replace-
ment levels would not immediately elimi-
nate population growth in a society.
This particular bank's egalitarian policies
toward women seem to have overcome bu-
reaucratic inertia because they were imple-
mented during a time of extraordinary or-
ganizational growth and change. When
analyzed in 1968, the bank was described
as one of the "largest and fastest growing
business concerns in the nation" (Narrative
report 415, 19681. Employment increased
nearly 50 percent between 1961 and 1968,
when the bank was automating its data pro-
cessing operation and establishing regional
offices and branch banks throughout south-
ern California. Growth and technological
change appear to be directly responsible for
the desegregation of several management,
administrative, and data processing job cias-
sifications in this firm. Its atypical experi-
ence demonstrates that the segregation ob-
served in other large establishments is not
inevitable. i9 These jobs, however, may have
,9 Other data gathered in California by the Employ-
ment Service corroborate this. They studied about 30
other establishments with more than 800 employees
not included in our analysis because of incomplete cov-
erage of some aspect of their operations. Of these, only
3 were moderately desegregated: a university campus
(/\ = 70.5), a unified school district (/\ = 76.5), and
an insurance company (/\ = 77.21. Among large bu-
reaucracies, assignment of both sexes to the same job
titles occurs most often in social overhead capital or-
ganizations- firms in health, education, and welfare
services, transportation, utilities, finance, insurance,
and real estate industries. But the banking and finance
industry is not uniformly desegregated. Of the 5 other
establishments in our sample engaged in such activi-
ties, 3 were completely segregated, 1 was nearly so At\
= 93.3), and 1 was moderately integrated (I\ = 73.7).
OCR for page 48
48
WILLIAM T. BIELBY AND [AMES N. BARON
TABLE 3-7 Descriptive Statistics for Longitudinal Sample (N = 75)
Time 1
Time 2
Standard Standard
Variable Mean Deviation Mean Deviation
Log size 4.26 1.21 4.38 1.21
Log specialization 1.37 0.99 1.48 1.03
Manufacturing industry .79
Core sector .61
Ambiguous sector .32
Periphery sector .07 - —
Log fragmentation .18 .20 .17 .19
Union/bidding arrangements .37 — .39
Proportion women .29 .2S .29 .24
Proportion production workers .64 .30 .62 .29
Proportion clerical and sales workers .16 .17 .18 .18
Proportion service workers .08 .22 .09 .22
Proportion PI Ma workers .10 .12 .11 .1 1
a Professional, technical, and managerial.
been integrated precisely because they were
new; it may take several years for the sex
label of a new line of work to become es-
tablished.
Longitudinal Analyses: The Permanence of
Sex Segregation
Seventy-fve of the 393 establishments in
our sample were studied more than once.
The average interval between visits was about
5 years, with a range of 2 to 12 years. The
size composition of the follow-up sample is
very consistent with the age and size-spe-
cif~c establishment mortality rates reported
by Birch (19791. While moderately large,
older enterprises were slightly more likely
to be revisited, we detected no other sys-
tematic biases in the Employment Service's
choice of establishments for follow-up anal-
yses (for details see Baron, 1982, Chapter
VII).
Descriptive statistics for the 75 establish-
ments are reported in Table 3 7.2° Estab-
20 Several establishments were studied more than twice
by the Employment Service. In those instances, we
selected the pair for which the interval between anal-
yses was closest to 5 years. Consequently, analyses for
16 of the establishments in the longitudinal analyses
do not include the one selected for the cross-sectional
sample of 393 observations.
lishments selected for restudy were slightly
larger on average and more concentrated in
manufacturing; other differences between
descriptive statistics for these observations
and those for the entire sample reflect con-
comitants of organizational scale (cf. Tables
3-1 and 3-71. In most instances, the temporal
data describe changes between the mid-1960s
and the early 1970s the period immedi-
ately following passage of Title VII of the
Civil Rights Act of 1964. Most establish-
ments expander} employment between anal-
yses; the labor force was stable in just 6 cases
and was reduced in 25 establishments.
As Table 3-8 shows, neither legislation
nor organizational change effected much
change in sex segregation in the late 1960s
and early 1970s. Two-thirds of the estab-
lishments remained all-male, completely
segregated, or almost fully segregated (see
the 9 cells in the bottom-right corner of the
table). Indeed, much of the change in Table
3-8 reflects very small differences in seg-
regation indices.
Table 3-9 lists characteristics of the 18
establishments for which the segregation in-
dex changed by at least 5 points. Only a few
attributes differed systematically between
the 11 organizations that became less seg-
regated, between the 7 that became more
segregated, and between the 57 in which
OCR for page 49
SEX SEGREGATION WITHIN ORGANIZATIONS
-
TABLE 3-8 Segregation Levels Over Time (N = 75)
Segregation: Time 2
Segregation: Time 1
Moderate (lt ~ 75)
Very
Mod High High
All
Complete Male Total
High (75 < ~ < 90)
Very high (90 < ~ < 100)
Complete Hi= 100)
All male
TOTAL
3 2 2 2
4 15 3
4 5 29 2
l
5 10 23 34 3
(7%) (13%) (29%) (47%) (4%)
(1%)
9
(12%)
23
(31%)
40
(53%)
2
(3%)
75
(100%)
sex segregation remained virtually constant.
Unionized establishments and those with
formal bidding arrangements tended not to
change from their high levels of segregation,
nor did those in which women comprised a
small minority of the work force. Organi-
zations that partially desegregated were typ-
ically in small-scale manufacturing or were
providers of personal and social services.
Most declines in segregation are attrib-
utable to a nominal change in the compo-
sition of 1 or 2 jobs rather than a major
change in personnel practices. For example,
2 male checkers were hired by a large-chain
supermarket, a mate operator was hired by
a firm providing mobile telephone service,
and a woman was hired to work as a lens
finisher in a firm manufacturing contact len-
ses. The "integrated" title of chip girl in the
poker cardroom is completely segregated by
shift; apparently, only males work in the
early morning hours. 2} A company that man-
ufactured silkscreened wall coverings hired
4 female inspectors, but continued to dis-
criminate statistically, hiring women only as
inspectors, paint mixers, and clericals "due
to occasional job requirements of lifting heavy
21 The apparent integration of the chip girl position
may simply reflect a typographical error on the 1970
staffing schedule. If so, the establishment was perfectly
segregated in 1966 and 1970.
reels of paper" in the other production work
(Narrative report 1476, 19701. Only a man-
ufacturer of kitchen ranges displays a delib-
erate effort to desegregate its work force.
Only 2 women were employed in produc-
tion departments in 1966. As in the case of
the bank described earlier, organizational
growth appears to have facilitated gender
parity. By 1970 the work force of this man-
ufacturing firm had more than doubled, and
women worked alongsicle men in 14 entry-
level jobs in the assembly department, re-
clucing the segregation index from 99.5 to
78.8. Nevertheless, as in the bank, employ-
ment in several other nonadministrative de-
partments remained restricted to males in
1970.
The 7 establishments that became more
segregated over time were all highly seg-
regatecT initially, ant! typically the segrega-
tion of only 1 job produced most or all of
the change. In several instances, increased
segregation took place in the context of or-
ganizational retrenchment or consolidation.
But in no case was there any evidence that
increased segregation resulted from clelib-
erate employer strategy to manipulate the
sex composition of jobs. Given the high ini-
tial levels of segregation, the small increases
in the index could easily be attributable to
random perturbations in labor supply and
demand.
OCR for page 50
so
WILLIAM T. BIELBY AND JAMES N. BARON
TABLE 3-9 Attributes of Establishments With Increasing or Decreasing Levels of Sex
Segregation
Number of Number of Segregation
Entry Year Employees Females Index
No. Products or Activities T1-T2 Sectora Industryb T1-T2 T1-T2 T1-T2
Establishments With Decreasing Segregation
1 Supermarkets 1961-1967 A Trade 56-49 16-12 100.0-94.6
2 Optical goods 1965-1968 C Manufacturing 42-44 18-21 77.8-69.6
3 Mental health facilities 1966-1970 A SOC 149-145 124-118 96.0-86.4
4 Wall coverings 1965-1970 C Manufacturing 139-142 12-26 96.1-85.8
5 Mobile telephone service 1966-1971 P SOC 15-30 13-21 100.0-88.9
6 Clay products 1964-1971 A Manufacturing 20-35 3-11 100.0-87.5
7 Poker cardrooms 1966-1970 A Services 149-190 97-66 100.0-87.1
8 Optical goods 1965-1968 A Manufacturing 9-10 3-3 100.0-85.7
9 Kitchen ranges 1966-1970 C Manufacturing 198-438 8-32 99.5-78.8
10 Wigs 1965-1971 A Manufacturing 11-11 10-6 90.0-63.3
11 Wall coverings 1965-1970 C Manufacturing 27-54 6-17 95.2-67.9
Establishments With Increasing Segregation
12 Trophies 1961-1971 C Manufacturing 75-58 29-33 86.6-92.0
13 Electronic instruments 1967-1970 C Manufacturing 95-49 20-11 84.0-89.5
14 Pottery 1960-1968 C Manufacturing 309-535 137-294 92.7-99.2
15 Musical instruments 1960-1970 C Manufacturing 164-173 10-4 89.0-95.9
16 Ceramic tiles 1971-1977 A Manufacturing 21-42 9-18 91.7-100.0
17 Securities exchange 1960-1968 A SOC 60-55 42-21 80.1-100.0
18 Garments 1960-1970 A Manufacturing 25-26 16-17 77.8-100.0
a p = periphery; A = ambiguous; C = core.
b S0c = social overhead capital.
In sum, changes in organizational forms
and environments and shifting labor supply
and demand had little eject on sex segre-
gation in the late 1960s and early 1970s.22
In most of the establishments we examined,
the consistently high levels of segregation
are probably due to long-standing policies
for hiring and allocating workers, perhaps
reflecting industrywicle practices predating
the establishment itself. Neither a changing
22 Nineteen establishments were visited 3 or more
times by the Employment Service. Seven were per-
fectly segregated at each visit, and segregation de-
creased between analyses in only 1 establishment. In
short, there is no evidence of a secular trend in job
segregation by sex between 1959 and 1979.
political and legal climate nor an influx of
women workers in the late 1960s affected
definitions of women's work within this sam-
ple of work organizations, especially in large,
unionized establishments with institution-
alized procedures governing hiring and ad-
vancement.
DISCUSSION
In most establishments, few job ciassifi-
cations are staked by both men and women.
IncleecI, complete segregation was the norm
in establishments studied in California by
the U. S. Employment Service between 1959
and 1979, and segregation levels were vir-
tually constant in these organizations during
OCR for page 51
SEX SEGREGATION WITHIN ORGANIZATIONS
5
TABLE 3-9 (Continued)
If segregation due mostly to one title:
T1 T2
Number Number
Number of Number of
of Males Females of Males Females Comments
Entry
No. Title
1
2
3
4
5
6
7
8
9
Checker
Several
Several
Several
Operator
Press operator
Chip girl
Finisher
Several
10 Ventilator
11 Several
12 Several
13 Assembler
14 Packer
15 Assembler
16 Kiln loader
17 Tabulation clerk
18
Establishments With Decreasing Segregation
0 10 2
0 10 1
2 0 3
0 35 16
2 0 1
10
5
3
1
6
Sewing machine
operator
1
13
5
3
1
5
3
9
2
18
1
1
3
Establishments With Increasing Segregation
3
o
5
o
8
8
o
o
o
3
5 Managerial promotion track for box boy but not for
checker.
Integration of 2 lens grinding jobs.
Integration of kitchen helper (initially male) and head
nurse (initially female) titles.
Integration of 2 inspector titles (initially male).
Chip girl title segregated by shifts.
Desegregated 14 of 19 jobs in assembly department; 7 of
8 other departments completely segregated.
Integration of 1 nonproduction and 3 production jobs; 23
of 28 jobs still segregated in 1970.
Discontinued jewelry manufacturing, eliminating 2
integrated jobs.
Decline in production between 1967 and 1970.
Kiln loader only integrated job in 1971.
Data processing operation consolidated as separate
subsidiary between 1960 and 1968.
Administrative jobs not included in staffing schedule for
- 1970.
the late 1960s and early 1970s. When bal-
anced sex ratios did occur, they almost al-
ways reflected just l or 2 integrated job titles
within an establishment.
Multivariate analyses revealed that or-
ganizational scale is strongly associated with
levels of sex segregation. Small enterprises
are either completely segregated or trivially
desegregated; larger enterprises tent! to have
almost all workers in segregated jobs. Union
contracts and formal bidding procedures,
positional specialization, reliance on trained
nonentry personnel, and manual job tasks
facilitate a division of labor segregated by
sex, while women are less extensively seg-
regated when they comprise more than a
small minority of the work force.
Institutional arrangements that prevail in
core firms rather than sectoral location per
se shape sex segregation, and they are
most visible in the largest establishments in
our sample. Many had sex-specific policies
for allocating workers, oftentimes sanc-
tione(1 by collective bargaining agreements.
In manufacturing establishments, legal re-
strictions on physical demands of women's
work typically rationalized complete sex seg-
regation among production jobs, even when
lifting was only required infrequently. Estab-
lishments in the economic core also specialize
skills and fragment job tasks more than other
enterprises, and these aspects of bureaucracy
are not gender-neutral.
Of course, most women are not employed
OCR for page 52
52
WILLIAM T. BIELBY AND [AMES N. BARON
in such manufacturing contexts. Less than
one-quarter of all women workers are in
manufacturing enterprises with 50 or more
employees, and one-half the females in non-
manufacturing industries work in establish-
ments with fewer than 50 employees.23 Even
small organizations outside of manufactur-
ing, however, are highly segregated. Of the
100 nonmanufacturing establishments in our
sample having fewer than 50 employees, 73
were perfectly segregated, and women em-
ployed in those establishments are more likely
to be segregated through the patriarchical
actions of an entrepreneur or supervisor than
by institutionalized bureaucratic rules and
procedures. In short, there is equifinality in
sex segregation of the workplace. Personal
and patriarchal control prevails in small es-
tablishments, bureaucratic structures seg-
regate men and women in large nonmanu-
facturing enterprises, and technical control
excludes women from many production jobs
within the manufacturing sector.
Do these findings imply that sex segre-
gation is immutable? We detected little
change in segregation indices, even among
firms analyzed after court decisions in 1971
struck down California's restrictions on
women's lifting. 24 In 1968, affirmative action
goals and timetables were required of firms
holding federal contracts, yet establish-
ments in our sample covered by this order
23 Figures are based on the 1967 Technological Ad-
vance in an Expanding Economy survey by the Uni-
versity of Michigan's Institute for Social Research
(Mueller et al., 1969~. Of 766 women working outside
the home, 560 were employed in nonmanufacturing
industries, and 282 of those were in establishments
with fewer than 50 employees. Since then, the female
work force has become even less concentrated in man-
ufacturing, although there may have been a shift to-
ward employment in larger establishments.
24 Of 46 establishments analyzed between 1972 and
1979, 26 were completely segregated. Eleven had seg-
regation indices less than 75, but 5 were real estate
enterprises studied by the Employment Service in 1973.
Levels of segregation remained uniformly high in man-
ufacturing establishments studied after 1971.
were no less segregated than others (see also
Salancik, 19791.
Nevertheless, we can point to some ex-
traordinary circumstances under which de-
segregation did occur. As noted above, one
bank and a manufacturing enterprise inte-
grated a number of nonclerical jobs during
a time of rapid growth or technological change
or both, before there was any substantial
government pressure to change personnel
policies regarding minorities. Short ofthese
fortuitous and idiosyncratic circumstances,
large and systematic reductions in gender
segregation seem unlikely to occur in the
absence of fundamental shocks to the social
system. For example, during World War II,
employers faced extreme labor shortages.
California's Industrial Welfare Commission
granted 6Q permits that exempted women
from its 25-pound restriction on lifting, al-
lowing 4,539 women workers to enter pro-
duction jobs formerly closed to them.25 Thus,
unusual extraorganizational circumstances
forced employers to reject long-standing
practices based ostensibly on physiological
differences between the sexes. Neverthe-
less, after the war these women were de-
mobilized as rapidly as they had been in-
tegrated into the work force. The procedures
for obtaining exemptions from the state la-
bor cocle remained in effect but were barely
utilized, and until 1971 the code rational-
ized sex segregation of production jobs in
manufacturing.
To the extent that our results are gener-
alizable, two policy implications of our study
are clear. First, sex segregation will un-
doubtedly persist if policy makers adopt a
laissez faire stance. Neither demographic
trends, nor technological changes, nor bu-
reaucratic imperatives are natural forces Mat
lead to balanced sex ratios within jobs or
firms. Second, policy intervention is un-
25 Personal communication, Margaret Miller, Exec-
utive Officer, California Industrial Welfare Commis-
sion.
OCR for page 53
SEX SEGREGATION WITHIN ORGANIZATIONS
.
53
likely to make matters worse most estab-
lishments are about as segregated as they
can possibly be. While there may be polit-
ical or economic incentives for focusing on
certain kinds of work contexts, policy efforts
could justifiably be directed at almost any
area of economic activity, since almost all
establishments are equally segregated.
Our findings also suggest some strategies
for attacking sex segregation. First, policies
that segregate men and women are most
visible in large organizations. They are doc-
umented in written contracts, rules, job def-
initions, and procedures; they do not exist
solely in subjective tastes of employers, em-
ployees, and clients. Large firms are often
subject to public scrutiny, and their greater
dependence on government contracts and
regulations makes them susceptible to pol-
icy interventions (Salancik, 19791. They are
also more likely to have slack resources with
which to absorb the costs of social change.
In short, it seems reasonable first to redress
visible and easily identifiable mechanisms
of sex segregation in organizations that are
vulnerable to outside pressures, even if only
a fraction of all women work in such estab-
lishments.
Second, interventions seem more likely
to succeed in organizations that already have
a sizable female work force. Sex ratios affect
the balance of power among organizational
constituents (Kanter, 1977a), and our results
show that, as women comprise a larger per-
centage of an organization's labor force, em-
ployers seem less likely to segregate them.
Changing the sex composition of jobs will
require modifying organizations' rules for
advancement through internal labor mar-
kets. Such changes are easier to accomplish
when female workers command firm-spe-
cific experience. In short, segregation seems
likely to persist in the absence of severe
external pressures on the organization. Fur-
thermore, the technical and political viabil-
ity of efforts to abate the sexual division of
labor depends on the existence of a constit-
uency inside. The presence of a visible con-
tingent of minority employees within an or-
ganization even at the lowest ranks may
facilitate efforts to desegregate work.
Efforts to impose workplace equity in the
absence of such a constituency might ac-
tually backfire. "Tokens" or "solos" can so-
lidify resistance by male workers and de-
moralize those who should champion and
benefit from equal employment opportunity
programs (Kanter, 1977a; NortheraPc and
Martin, 1981~. This underscores once again
the limits of laissez faire approaches, since
nearly all of the natural desegregation we
observed in our sample involved tokenism.
Our recommendations are based on statis-
tical associations rather than examinations of
specific policy interventions. Nevertheless,
recent surveys of attempts to reduce sex seg-
regation do document the effectiveness of the
organizational strategies we have suggested
(Shaeffer and Lynton, 1979; O'FarreD and
Harlan, 19821. One conclusion cannot be dis-
puted: Doing nothing guarantees persistent
sex segregation in all areas of economic activ-
ity. Although not the focus of this study, we
repeatedly encountered instances of sex seg-
regation of jobs leading to gender-specific pro-
motion lines an orderly progression through
jobs of successively greater authority and re-
sponsibility for men and dead-end careers for
women. The pervasive sex segregation across
organizational and institutional contexts that
our study has documented almost certainly
accounts for the substantial sex differences re-
vealed by individual-level analyses of work
inequalities (e.g., Baron and Bielby, 1982;
Bielby and Baron, 1982; Wolf and Fligstein,
19791.
The degree and persistence of sex seg-
regation leave us somewhat pessimistic about
prospects for rapid or extensive change. On
the other hand, by documenting the impact
of organizational structures and dynamics on
the sexual division of labor ant! by un-
derscoring the pervasiveness of gender seg-
regation we hope to facilitate more in-
formed research efforts and policy inter-
ventions in the future.
OCR for page 54
54
_ WILLIAM T. BIELBY AND [AMES N. BARON
ACKNOWLEDGMENTS
The authors were supported in part by a
grant from the National Science Foundation
(SES 79-24905~. We gratefully acknowledge
the cooperation and assistance of the Cali-
fornia Occupational Analysis Field Center.
Our thanks go to Isabelle M. Aliain, Hallie
Anderson, and Kelly Moreno who assisted
in this research, and to Trivia Miller who
processed many versions of this manuscript.
We are indebted to our many colleagues
who have influenced our thinking on sex
segregation and have commented on earlier
versions of this report, especially to Donald
I. Treiman and Karen O. Mason for their
detailed suggestions.
REFERENCES
Arrow, Kenneth
1973 "The theory of discrimination," Pp. 3-33 in Or-
ley Ashenfelter and Albert Bees (eds.), Dis-
crimination in Labor Markets. Princeton, N.J.:
Princeton University Press.
Averitt, Robert
1968 The Dual Economy. New York: Horton.
Baron, James
1982 Economic Segmentation and the Organization
of Work. Unpublished Ph.D. dissertation,
University of California, Santa Barbara.
Baron, James, and William Bielby
1982 "Workers and machines: dimensions and de-
terminants of technical relations in the work-
place." American Sociological Review 47:175-
188.
Bielby, Denise
1978 "Career sex-atypicality and career involvement
of college educated women: baseline evidence
from the 1960's." Sociology of Education 51
Qanuary):7-28.
Bielby, William, and James Baron
1982 "Organizations, technology, and worker attach-
ment to the firm." In Donald J. Treiman and
Robert V. Robinson feds.), Research in Social
Stratification and Mobility, Vol. II. Greenwich,
Conn.: JAI Press.
Birch, David
1979 The Job Generation Process. Cambridge, Mass.:
MIT Program on Neighborhood and Regional
Change.
Blau, Francine
1977 Equal Pay in the Office. Lexington, Mass.:
Lexington Books.
Blau, Francine, and Wallace Hendricks
1978 "Occupational segregation by sex: trends and
prospects." Journal of Human Resources 14
(2~:197-210.
Blau, Francine, and Carol Jusenius
1976 "Economists' approaches to sex segregation in
the labor market: an appraisal." Pp. 181-99 in
Martha Blaxall and Barbara Reagan (eds.),
Women and the Workplace: The Implications
of Occupational Segregation. Chicago: Uni-
versity of Chicago Press.
Braverman, Harry
1974 Labor and Monopoly Capital. New York:
Monthly Review Press.
Bridges, William, and Richard Berk
1974 "Determinants of white-collar income: an eval-
uation of equal pay for equal work." Social Sci-
ence Research 3:211-34.
Doeringer, Peter, and Michael Piore
1971 Internal Labor Markets and Manpower Anal-
ysis. Lexington, Mass.: D. C. Heath.
Duncan, Otis, and Beverly Duncan
1955 "A methodological analysis of segregation in-
dexes." American Sociological Review 20:210-
17.
Edwards, Richard
1979 Contested Terrain. The Transformation of the
Workplace in America. New York: Basic Books.
England, Paula
1981a "Assessing trends in occupational sex segre-
gation, 1900-1976." Pp. 273-95 in Ivar Berg
(ed.), Sociological Perspectives on Labor Mar-
kets. New York: Academic Press.
1981b "Explanations of occupational sex segregation:
an interdisciplinary review." Unpublished
manuscript, University of Texas at Dallas (Jan-
uary).
1982 "The failure of human capital theory to explain
occupational sex segregation." Journal of Hu-
man Resources 17:358-70.
Gibbs, Jack, and Dudley Poston, Jr.
1975 "The division of labor: conceptualization and
related measures." Social Forces 53:468-76.
Gross, Edward
1968 "Plus ca change . . . ? The sexual structure of
occupations over time." Social Problems 16:198-
208.
Halaby, Charles
1979 "Job-specific sex differences in organizational
reward attainment: wage discrimination vs. rank
segregation." Social Forces 58 (September):
108-127.
Kanter, Rosabeth
1977a "Some effects of proportions on group life:
skewed sex ratios and responses to token
women." American Journal of Sociology 82:965-
90.
1977b Men and Women of the Colporation. New York:
Harper & Row.
Marini, Margaret
1980 "Sex differences in the process of occupational
attainment: a closer look." Social Science Re-
search 9:307-61.
OCR for page 55
SEX SEGREGATION WITHIN ORGANIZATIONS
55
1978 "Sex differences in the determination of ado-
lescent aspirations: a review of research." Sex
Roles 4:723-53.
Milkman, Ruth
1980 "Organizing the sexual division of labor: his-
torical perspectives on women's work and the
American labor movement." Socialist Review
49:95-150.
Miller, Ann R., Donald J. Treiman, Pamela S. Cain,
and Patricia A. Boos
1980 Work, Jobs, and Occupations: A Critical Re-
view of the Dictionary of Occupational Titles.
Washington D.C.: National Academy Press.
Mueller, Eva, Judith Hybels, Jay Schmiedeskamp, John
Sonquist, and Charles Staelin
1969 Technological Advance in an Expanding Econ-
omy. Ann Arbor, Mich.: Institute for Social
Research, University of Michigan.
Northcraft, Gregory, and Joanne Martin
1981 "Double jeopardy: why some women and mi-
norities~object to affirmative action programs."
Research paper 599, Graduate School of Busi-
ness, Stanford University.
O'Farrell, Brigid, and Sharon Harlan
1982 "Job integration strategies: today's programs
and tomorrow's needs." Paper presented at the
National Research Council's Workshop on Job
Segregation by Sex, May, Washington, D.C.
Polachek, Solomon
1979 "Occupational segregation among women: the-
ory, evidence, and a prognosis." Pp. 137-57 in
Cynthia B. Lloyd, Emily S. Andrews, and Cur-
tis L. Gilroy feds.), Women in the Labor Mar-
ket. New York: Columbia University Press.
Salancik, Gerald
1979 "Interorganizational dependence and respon-
siveness to affirmative action: the case of women
and defense contractors." Academy of Man-
agement Journal 22:375-94.
Shaeffer, Ruth Gilbert, and Edith F. Lynton
1979 Corporate Experiences in Improving Women's
Job Opportunities. New York: The Conference
Board.
Simmel, George
1950 "Quantitative aspects of the group." Pp. 87-
177 in Kurt H. Wolff (ed.), The Sociology of
George Simmel. New York: Free Press. Orig-
inally published in 1923.
Thomas, Robert
1980 "The social organization of industrial agricul-
ture." Paper presented at the 1980 meeting of
the American Sociological Association, New
York.
Treiman, Donald J., and Heidi I. Hartmann feds.)
1981 Women, Work and Wages: Equal Pay for Jobs
of Equal Value. Washington, D.C.: National
Academy Press.
Tyack, David, and Myra Strober
1981 "Jobs and gender: a history of the structuring
of educational employment by sex." Pp. 131-
52 in Patricia Schmuck and W. W. Charters
feds.), Educational Policy and Management:
Sex Differentials. New York: Academic Press.
U.S. Bureau of the Census
1976 County Business Patterns, United States Sum-
mary. Washington, D.C.: U.S. Government
Printing Office.
U.S. Department of Labor
1972 Handbook for Analyzing Jobs. Manpower
Administration. Washington, D.C.: U.S. Gov-
ernment Printing Office.
Williams, Gregory
1979 "1he changing U.S. labor force and occupa-
ffonal differentiation by sex." Demography
16:73-88.
Wolf, Wendy C., and Neil D. Fligstein
1979 "Sex and authority in the workplace: the causes
of sexual inequality." American Sociological
Review 44:232-52.
Representative terms from entire chapter:
employment service
