Skip to main content

Full text of "ERIC ED056332: Women in the Occupational World."

See other formats


ED 056 332 

CS 006 709 





Levitin* Teresa 

Women in the Occupational World. 

Michigan Univ. * Ann Arbor. Survey Research Center. 
Sep 71 

lip.; Paper presented at American Psychological 
Association Annual convention (79th* Washington* D. 
C.* September 3-7* 1971) 

MF-$0 • 65 HC-S3.29 

^Discriminatory Attitudes (Social) ; Employment; 
^Females; ^income; Jobs; Job Satisfaction; Labor 
Force; Salary Differentials; ^Working Women 


The purpose of this paper is to demonstrate that 
women do not receive occupational rewards commensurate with their 
achievement* rewards that are allocated to equally qualified men. The 
analysis of discrimination is directed toward 3 problems: (1) to what 
extent are women denied occupational rewards that* according to 
achievement ideology* they have legitimately earned; (2) what are the 
demographic and occupational distributions of reward inequalities 
among working women; and (3) to what extent are the researchers 
objective measures of discrimination associated with reports of 
perceived discrimination. Data were obtained from a survey of 
American workers conducted by the survey Research center of the 
University of Michigan late in 1969. Results showed that the average 
working woman received $3*458 less than her male counterpart. In 
regard to demographic and occupational variables that are related to 
severity of discrimination* it was found that the women who lost 
$3*500 or more were the youngest (16-29 years old) and the oldest (55 
years plus) • (Author/RK) 











Women in the Occupational World*'" 

Teresa Levitin 
Survey Research Center 


The University of Michigan 

Paying women less than men is easily justified. Justifi- 
cations vary, but they often include the claim that women are 
more likely to be sick, to be absentees, and to quit their 
jobs. Aside from the factual errors in these claims 
(Dornbusch, 1966), the form of the argument illustrates a 
critical but often neglected point! discrimination is usually 
justified by reference to a particular ideology or set of 

values. Most often, an achievement ideology is used to justify 
differential payment, and the argument goes as follows: 
unequal pay is legitimate when there is unequal achievement. 
Since women achieve less, it is legitimate, not discriminatory, 
to pay them less* Of course* adherence to this ideology 
should mean that women who achieve as much as men receive 
equal occupational rewards. 

The purpose of this paper is to demonstrate that women 
do no t receive occupational rewards commensurate with their 
achievement, rewards that are allocated to equally qualified 
men. Rather, women experience occupational discrimination, 
discrimination being defined here as the provision of fewer 
rewards or facilities than are legitimately deserved. Although 
occupational sex discrimination, particularly among professional 
women, has been well documented (Austin, 1969; Epstein, 1970; 

1 ^ 

Paper presented at the 79th Annual American Psychological 

Association Convention, Washington, D.C., September 3-7, 1971. 


Mattfeld and Van Aken r 1965), this study is unique in 
the data are drawn from a national probability sample of 
currently employed workers, female and male, in a variety of 
occupations. The findings are thus general izable to the 
population of American working women and men. Still another 
feature of this research is that we have developed measures 
to detect individual differences in the amount of discrim- 
ination encountered. Discrimination is usually inferred by 
examining group differences. For example, evidence of 
discrimination against blacks is gathered by comparing occu- 
pational or economic differences between blacks and whites 
(Batchelder, 1968; Kahn, 1968; Ross, 1967). With the data 
reported in this paper, we are able to carry out more inten- 
sive studies to assess the conditions associated with 
different degrees of discrimination. 

Our analysis of discrimination is directed toward three 
problems i first, to vhat extent are women denied occupational 
rewards that, according to an achievement, ideology, they have 
legitimately earned? Specification of some ideology or set 
of values is necessary to distinguish legitimate from discrim- 
inatory differentiation. In our society, and particularly in 
the economic domain, the achievement ideology is dominant and 
pervasive. We therefore chose it as the framework within 
which to study occupational sex discrimination, a choice that 
does rot necessarily indicate adherence to this set of values. 
There are other, alternative ideologies. For example, Marxist 
writings describe a need ideology,,, according to which rewards 
ought to be based on need, rather than on performance. The 




second research problem was to discover the demographic and 
occupational distributions of reward inequalities among working 
women. Third, we explored the extent to which our objective 
measures of discrimination were associated with reports of 
perceived discrimination, A more complete statement of the 
research problems, the method and its limitations, and our 
results can be found in a forthcoming article by Levitin, 

Quinn and Staines, 

Data were obtained from a survey of American workers 
conducted by the Survey Research Center of the University of 
Michigan late in 1969. The research was supported by a 
contract with the Employment Standards Administration of the 
U.S. Department of Labor. The interpretations and viewpoints 
presented here do not necessarily represent the official 
position or policy of the Department of Labor. 

The principle aims of the survey were to determine some 
of the problems workers face, to develop measures of job 
satisfaction and mental health, and to assess the effects of 
working conditions on both job satisfaction and mental health. 

Basic univariate and bivariate tables are available elsewhere 
(Quinn, et al. , 1970), This analysis represents a preliminary 
part of a forthcoming report on the status of working women. 
The sample was a national probability sample of persons 
who were living in households, were 16 years old or o3der, and 
were working for pay 20 hours a week or more. Unemployed 
members of the labor force and those outside the labor force 
were thus excluded. All eligible workers were interviewed in 

each of the sampled households, and every worker in the 
population had an equal probability of being selected. The 





full sample included 539 women and 933 men. For the present 
analysis, three groups of workers were excluded: self- 

employed workers, part-time workers, defined as those working 
less than 35 hours a week, and workers who were seasonally 
or irregularly employed during the year. School teachers 
were not regarded as being irregularly employed. The remain- 
ing sample consisted of 351 women and 695 men. For some 
analyses, the sample of men was further randomly divided into 
two ha If- samples . 

Following an analysis begun last year by Graham Staines,- 
Rob Quinn, Graham Staines and I chose two measures of occu- 
pational rewards: total annual income from the worker’s 

primary job before taxes or other deductions and the overall 
quality of working conditions. The latter is measured by a 
Quality of Work Index, a summary index assessing such areas 
as income, health and safety, work hours, transportation to 
and from work, interpersonal relations on the job, job 
security, and the content of the worker’s job. Details of how 
this and other indices were constructed, tested for reliability, 
and utilized may be found in the already mentioned report of 
univariate and bivariate tables (Quinn, et al. , 1970) and in 

forthcoming reports. 

We also chose several measures of merit or performance 
that, according to an achievement ideology, ought to predict 
to different levels of these two occupational rewards* We had 
no objective criteria to assess each person’s past or antici- 
pated performance, such as units produced per hour| nor, given 
the occupational heterogeneity of the sample, could we expect 
to have any. Thus, the six predictor variables chosen were 



only indirect indicators of performance or achievement. These 
six predictor variables were education, tenure with one's 
employer, tenure on one's specific job with that employer; 
number of hours worked each week; amount of supervisory 
responsibility; and occupational prestige as measured by the 
Duncan scale (Reiss, et al., 1951). The choice of appropriate 
indicators of achievement was somewhat arbitrary, reflecting 
our understanding of the achievement ideology. For example, 
whether or not job tenure is an appropriate factor for 
determining income depends upon how tenure is conceptualized. 
It was decided to assume that longer tenure may lead to the 
acquisition of additional experience and skills and thereby 
constitutes an appropriate basis for assigning occupational 

rewards , 

Objective sex discrimination was then defined as the 
difference between how much each woman was rewarded and how 
much she ought to have been rewarded based on her scores on 
the six achievement predictor variables. To obtain a measure 
of objective discrimination, the assumption was made that 

occupational sex discrimination was not operative for men. 
Multiple regressions were calculated on a random half-sample 
of men to determine the optimal weighting of these six achieve- 
ment variables in predicting both reward measures for men. 

The obtained weights were used to compute expected values on 
the two occupat ional reward measures for both the sample of 
women and the second random half-sample of men. The measure of 


objective discrimination was computed by substracting the 
expected value from the observed* value of an occupational 
reward for each respondant, wit;h scale units in dollars for 

- 6 - 



annual income and on a 1-5 scale for the Quality of Work Index. 

Although we expected to discover that a woman received 
less occupational reward than a man with identical scores on 
the achievement predictors, we were hardly prepared for the 
size of the discrepancy between observed and exp ec ted annual 
income. The average woman actually received $3,458 (SD=$2,200; 
N=323) less than her male count erpart — theref ore $3,458 less 
than she should have received. Another way to state this 
result is to note that the median woman would have to receive 
71 p er c en t more than her current income to make that income 
equivalent to a man with the same scores on the predictor 
variables • 

Figure 1 shows graphically the distribution of total 
annual income discrepancies for all women and the second 
random half-sample of men. Fifty and three-tenths percent of 
the women had annual income discrepancies ranging from $3,000 
through $5,999. The mean annual income of 94.9 percent of 
the women was less than the amount they should have received 
on the basis of the achievement criteria. 

Sex differences in the distribution of the discrepancy 

scores on the Quality of Work Index were less extreme than 
those based on annual income. Nevertheless, 55 percent of the 
women had scores lower than their predicted scores. This 
result indicated that the quality of women’s occupations was 
less than would have been expected from their achievement 
scores * 

To discover which demographic and occupational variables 
were related to severity of discrimination, we dichotomized the 



sample of women into those who lost $3,500 or more in income 
and those who lost less. Briefly* the sighlficant differences 
were that the women who lost $3*500 or more were the youngest 
(16-29 years old) and the oldest (55 years or more) of the 
respondents. They were also white collar workers ; those 
employed in professional* technical, managerie 1 , clerical, and 
sales work; those who did not belong to a union; and those* in 
comparatively small establishments where less than 500 
employees worked. However, there were no significant associa- 
tions between severity of discrimination in reference to the 
Quality of Work Index and any of the selected demographic 
and occupational variables. 

Clearly, almost all women were discriminated against with 
regard both to their income and to the quality of their jobs, 
but only 7.9 percent reported differential treatment when 
asked, "Do you feel in any way discriminated against on your 
job because you are a women?" Thus, our objective measure of 
discrimination was virtually unrelated to perceived or reported 

Time constraints permit only a brief discussion of these 
results. The data clearly point to large discrepancies 
between the occupational rewards women earned and the occu- 
pational rewards they received in comparison to rewards allo- 
cated to equally qualified men. 

The achievement ideology is simply not equally applied to 
women and men* with 94*9 percent of the women being underpaid 
for their skills and their performance, underpaid, on the 
average, $3,458 a year. 


The argument is often made that even though women earn 
less than equally qualified men , this difference is offset by 
the better working conditions supposedly enjoyed by women. 
Whether or not a woman would be willing to trade a mean loss 
in annual income of $3,500 for better working conditions, more 
fringe benefits or other non-economic aspects of a job is an 
empirical question not tested here. What is clear from these 
data, however , is the fact that women were not receiving better 
quality jobs than would have been predicted from their per- 
formance, While there was ample evidence of income discrimin- 
ation, there was no evidence of "compensating” favoritism in 
terms of the quality of the job. In fact, just the opposite 
was true; women not only received lower pay, they also 
had worse jobs than equally qualified men, worse than they 
ought to have had based on their achievement. 

We have only presented first order associations between 
discrimination and the demographic and occupational variables. 
The discrepancy scores reflect all forms of illegitimate or 
discriminatory differentiation, including age and race as well 
as sex discrimination. Thus, we are now preparing to analyze 
the data further to try to separate sex from other kinds of 

Most bemusing is the fact that only 7.9 percent of the 
women reported on-the-job discrimination. No satisfactory 
answer as to why this figure is so low suggests itself, but 
reasons may include the followings Women may not know what 
equivalently qualified men are payed elsewhere, especially in 
other occupations. They may atti^bute some of the disparity 
O to factors they regard as legitimate. They may compare 

ERiC o 










- 5 - 

themselves to other women rather than to men. They may believe 
that j in principle, women and men should receive unequal 
occupational rewards. They may attribute discrimination to 
factors such as age and race rather than to sex. Or, they 
may restrict the term discrimination to instances in which 
discrimination is consciously planned and executed by some 
organizational decision-maker. We have no adequate explan- 
ation. However, the lack of a relationship between perceived 
and objective discrimination may be quite time-bound. Many 
political and social action groups are helping women become 
increasingly conscious of the discrimination they face and 
less tolerant of inequitable treatment. 

Thus, it is quite possible that there will be a sub- 
stantial future increase in the number of women reporting 
discrimination on their j obs--discriminat ion that the present 
study indicates they have every right to report. 




- 10 - 


Austin, H, The woman doctorate in America; Origins, career, and family . 

New York? Russell Sage Foundation, 1969, 

Batchelder, A.B. Decline in the relative income of Negro mien, Xn L,A. 

Ferman, J.C. Kornbluh, & J,A. Miller (Eds,), Negroes and labs . Ann Arbor? 

The University of Michigan Press, 1968, 65-92, 

Dornbusch, S.M. Afterword. In E.E, Maccoby (Ed.), The development of Sex 

differences , Stanford, California? Stanford University Press, 1966. Pp. 204-219, 

Epstein, D.F, Encountering the male establishment? Sex-status limits on 
women's careers in the professions, American Journal of Sociology . 1970, 

75, 965-982, 

Kahn, T, The economics of inequality. In L,A. Ferman, J,L, Kornbluh, & J.A. 

Miller (Eds.), Negroes and lobs . Ann Arbor: The University of Michigan Press, 

1968. pp. 15-28, 

Levitin, T. , Quinn, R..P, , & Staines, 6,L, Discrimination against the American 
working woman, American Behavorial Scientist , 197 1, in press. 

Mattfeld, J, & Van Aken, C,G, (Eds,)# Women and the Scientific professions . 

Cambridge, Mass.: The MIT Press, 1965, 

Quinn, R.P. , Seashore, S,, Mangione, T,, Campbell, D., Staines, G.L. , & 

McCullough, M, Survey of working conditions , Washington, D.C,, U.S. 

Government Printing Office, 1971. 

Reiss, A, J, Jr., Duncan, 0,D, , Hatt, P.K, , & North, C.C. Occupations and 
social status . New York? Free Press of Glencoe, 1961, 

Ross, A, and Hill, H. ,£ds.), Employment, race, and poverty . New Y 0 rk: 

Harcourt, Brace and World, Inc,, 1967, 

1 ® 


Figure 1. 

Percentage Distribution of Total Annual Income 
Discrepancies for Women and Second Random Half- sample of l!hn 








*H U> 

o e> 
O 0 > 

<eT cwT 



Survey Research Center 
The University of Michigan