[EPI welcomes
reader
feedback.]
The Conversion of Interests to Principals: The
Case of Comparable Worth
Once established, bureaucracies tend to become active in
ways that maintain or even enhance their existence. Such
tendencies are evident regardless of whether the activities
are consistent with the original raison d'etre of the
bureaucracy. For example, the welfare bureaucracy has a
vested interest in the perpetuation if not the expansion of
poverty. If it appears that poverty according to one
definition is declining, the solution is breathtakingly
simple: "poverty" is redefined so that it appears that we
are losing the war against it.
Similar redefinitions of "discrimination" have
characterized the civil rights and the feminist movement.
Significantly, black leaders generally opposed quotas in the
years immediately following enactment of the Civil Rights
Act of 1964. Over time, however, their expectations
concerning the status of black workers were not fulfilled.
As this happened, black civil rights leaders began to
support instead of oppose quotas, even though assurances
against quotas had been emphasized by the supporters of the
Civil Rights Act when it was enacted. Civil rights
organizations which began to experience difficulty
identifying victims of discrimination stayed in business by
expanding the meaning of discrimination or the categories of
victims of it.
Similarly, although discrimination and segregation on the
basis of sex has been largely eliminated, the hoped-for
consequences have not materialized, or have not materialized
as rapidly as some feminist leaders would like.
Consequently, they have redefined "discrimination" and what
constitutes evidence of it. Instead of the elimination of
discrimination, they seek an economy in which there are no
substantial differences in the proportion of males and
females in various occupations. This change is widely
referred to as "comparable worth."
The main argument for comparable worth can be summarized
as follows. Historically, women have been confined to a
small number of low-paying jobs. Because better-paying jobs
were not open to women on a nondiscriminatory basis, women's
wages were depressed even more by the fact that employers
could always frustrate efforts to increase the compensation
for women's work. As supporters of comparable worth have
noted, the most important source of wage differentials
between men and women is the concentration of women in
low-paying jobs. To the extent, therefore, that this pattern
continues, women's work is undervalued. Although there is
some penetration of the better-paying jobs by women, the
pattern of sex-based discrimination continues. Because it is
unfair to incumbents in predominantly female occupations to
switch to some other field, the fair solution is to pay them
more for what they are doing to redress the historic
inequity in the salary structure. To its supporters,
therefore, "comparable worth" and "pay equity" are
synonymous. In their view, pay equity requires eliminating
the practice of paying women less than men for work that
requires comparable skill, effort, responsibility, and
working conditions. They emphasize, however, that the
elimination of the differentials should be done by raising
the wages of the underpaid, female-dominated jobs.
To get at the heart of the issue, let us raise this
question: To what extent, if any, are the differentials in
male/female wages the result of voluntary labor market
decisions by women? These decisions may be to drop out of
the labor market to have children, or to devote time to
raising children, or to devote more time to the roles of
wife and/or mother. Likewise, they may be decisions to move
when the spouse relocates; in such cases, it is usually but
not always the female who is disadvantaged occupationally
and/or educationally.
Although estimates vary, the foregoing nondiscriminatory
explanations clearly explain most of the differential; an
estimate by a member of the U.S. Commission on Civil Rights
is that all but eleven percent of the differentials can be
explained this way.1,
2 If discrimination against females
accounted for the entire remaining differentials--an outcome
which is very unlikely--it would still account for only 11
percent of the differential at most. The fact is, however,
that there is little, if any, reliable evidence that a
significant portion of male/female wage differentials is
caused by discrimination against females. The supporters of
comparable worth simply attribute the differentials that are
not explained by conventional analysis as the result of sex
discrimination.3 In view of the
legal prohibitions against sex discrimination, this appears
to be an extremely dubious basis for estimating the extent
of sex discrimination.
My argument here is not that discrimination against women
does not exist. It is that there is no credible evidence
that it accounts for a significant portion of the
male/female wage gap. Indeed, when one considers the wide
range of voluntary decisions which contribute to this gap,
the attribution of it to sex discrimination must be viewed
as intellectually dishonest. In January 1983, men working
full-time had been on the job an average of 5.1 years; the
average for full-time female workers was 3.3
years.4 Obviously, these figures
reflect the fact that women are less attached than men to
labor force participation. In light of this fact, it is not
surprising that women choose to interrupt their employment
more often. They invest less in education and training,
because their payback is less. They are less likely to
undergo long periods of training and hence to reap the
rewards thereof. A 1977 study concluded that differences in
turnover and job-specific training accounted for 67-100
percent of the overall male/female wage differentials within
occupations. Not surprisingly, the proportion of women
employed in an occupation varies inversely with the amount
of the on-the-job training
required.5 Much as some may
lament the fact, women are predominant in occupations where
quitting and entry are easy.
Another factor widely ignored in comparing male/female
wage differentials is the number of hours worked. Bureau of
Labor Statistics data usually refer to wages received by
"full-time" workers. "Full-time" is defined as thirty-five
hours or more per week. This definition leaves considerable
room for sex differentials in the amount of time worked.
Consider that in one year (1982), 24.0 percent of male
workers, but only 10.0 percent of female workers, worked
over forty hours a week. Indeed, differences in hours worked
accounted for 9.2 percent of the overall male/female wage
gap in 1982.6
Clearly, the progressive income tax also discourages
female participation in the labor force. Inasmuch as the
spouse's income is likely to be taxed at high rates, there
is a major disincentive for spouses (usually women in this
situation) to undergo extensive training or to seek
high-paying work. In short, the effect of the income tax is
to encourage women to enter and leave the labor force as a
response to fluctuations in the husband's
income.7 Here again, voluntary
decisions by female workers, not "discrimination" against
them, help to explain the male/female wage gap.
Consider briefly how comparable worth is implemented. For
this purpose, a job evaluation study is made. First, job
descriptions are established through questionnaires and
interviews. Next, each job description is assigned points,
typically for skill, effort, responsibility, and working
conditions. After the points are assigned, salaries are
compared to determine whether jobs that have a similar
number of points have similar salaries. If there is a
difference, the differences are analyzed to determine
whether they are due to gender or race.
Not surprisingly, this approach almost invariable
concludes that jobs which are predominantly female are
undervalued, that is, underpaid. We should not be surprised
by this outcome for the following reasons.
By law or agreement, the studies are frequently performed
by sympathizers of comparable worth. Anyone familiar with
what passes for "research" in higher education knows that a
great deal of research consists of justifying conclusions
reached independently of it.
It is unrealistic to suppose that these studies are
conducted independently of the desired result. Where
consultants conduct the study, it would be naive to suppose
that its impact upon their future employment is not a
factor.
If one believes, as the comparable worth supporters do,
that the mere existence of male/female differentials
reflects discrimination, it is difficult to see how a study
could do anything but assign more points to the jobs which
are predominantly female.
Despite assertions that comparable worth can be
implemented "objectively," job comparisons are not
"objective" merely because they are based on numerical
scores. Decisions about the criteria to be considered
involve an irreducible element of subjectivity. Likewise,
decisions on the relative weight given to factors of skill,
responsibility, and working conditions are a matter of
subjective judgment.
Let me illustrate the problem from a public school
perspective. Most teachers of physics and chemistry are
male, but most public school teachers are female. To my
knowledge, however, teacher unions never suggest that the
shortages of mathematics and science teachers justify
increasing the salaries of science teachers only; that is
not the way unions work.
The market demand for the kind of persons who can teach
mathematics and science is much greater than the demand for
elementary teachers. The shortages result from the
persistent neglect of market factors in teacher recruitment,
but comparable worth would render the problem even more
insoluble, if that is possible. First, the female employees
(and perhaps most of the male ones also) are not going to
concede that their jobs deserve a lower point score than
teaching physics or chemistry. And secondly, most teachers
are even less likely to accept, let alone urge, that market
factors be given some weight in salary determination. To
supporters of comparable worth, any effort to pay a group of
predominantly male teachers more than other teachers, who
are predominantly female, would only be compounding "job
segregation" and "pay inequity." This rhetoric should not
obscure the real damage resulting from neglect of market
factors under comparable worth.
In some respects, the implementation of comparable worth
would have consequences contrary to those intended by its
supporters. If, for example, the costs of female employment
are increased without any corresponding increase in
productivity, employers will inevitably try to reduce
employment in predominantly female occupations. Comparable
worth would also tend to freeze existing proportions of
males and females in existing occupations. Women would have
less incentive to transfer to preponderantly male
occupations, thus slowing down the trend to increasing
female entry into predominantly male occupations. It is also
likely that comparable worth will politicize decisions that
have hitherto been regarded as economic ones, to be made by
employers and employees at the workplace. With the advent of
comparable worth, the focus of attention will shift from
greater productivity to control of the wage board. This
would be a severe setback to efforts to increase the
productivity of our labor force.
Although comparable worth is hardly the wave of the
future, a few political leaders continue to support it. In
their personal calculus, the political support it generates
outweighs any loss of support resulting from its
implementation. Similarly, we can expect unions that are
predominantly female, such as K-12 education, to bargain for
comparable worth. Such efforts may be successful from time
to time. They do not prove that comparable worth is sound or
equitable public policy. Instead, such incidents merely
provide additional examples of Pareto's observation that men
(and women!) find it easy to convert their interests into
principles.
Footnotes
1 U.S. Commission on Civil
Rights, Comparable Worth: Issue for the 80's, Vol. 2,
Proceedings, p. 131, June 6-7, 1984, Washington, D.C., June
1985.
2 C.W. Baird, Comparable
Worth: The Labor Theory of Value and Worse, Government Union
Review, pp. 1-29, Winter 1985.
3 See, for example, the
testimony supporting comparable worth by former Secretary of
Labor, Ray Marshall in (1, pp. 119-121). There is virtually
no controversy over the fact that supporters of comparable
worth typically use the methodology cited to estimate the
extent of sex discrimination.
4 Monthly Labor Review, p. 24,
June 1984.
5 E.M. Landes, Sex-Differences
in Wages and Employment: A Test of the Specific Capital
Hypothesis, Economic Inquiry, p. 523, October 1977.
6 Monthly Labor Review, p. 25,
June 1984.
7 P.G. Germanis, Comparable
Worth--Part I: A Theory With No Facts, Heritage Foundation
Backgrounder, pp. 8-9, March 2, 1984.
|
Building a Competitive Education
Industry
