CAE Working Paper #03-07
The Economics and Law of Sexual Harassment in the Workplace

by

Kaushik Basu
June 2003
.

---

The Economics and Law of Sexual Harassment in the Workplace




Kaushik Basu





Kaushik Basu is Professor of Economics and C. Marks Professor of International Studies,
and Director of the Program on Comparative Economic Development, Cornell
University, Ithaca, New York. His email address is <kb40@cornell.edu> and his website
is http://people.cornell.edu/pages/kb40/.

---

2



Some years ago, a marshland (subsequently named Salt Lake) adjoining the city
of Calcutta was developed by the local government and sold as small plots, at a
subsidized price, to people who were not rich and who might not otherwise have been
able to afford their own property. Politicians masterminding the plan worried that, unless
some special precautions were taken, the rich would soon buy the land from the original
owners, thereby “depriving” them of their plots. Thus, a law was announced which
prohibited the sale of land by the original owners in Salt Lake. When I tell economists
about this law, they usually laugh at the folly of politicians. If an owner wants to sell
land, it must be that the owner expects to be better off by doing so; it is thus hardly a
favor not to allow the sale.

However, economists do not laugh when it is pointed out that under current U.S.
law, a firm cannot offer a job contract in which the pay is high and the benefits good --
but the employer reserves the right to sexually harass the worker. The fact that the worker
who accepts this job must find the cost of sexual harassment to be less than the benefits
associated with the job does not seem reason enough to allow such a contract. Though in
some ways the two examples may appear similar--both involve two adults choosing to
make an exchange that seems to have no obvious negative externality on others— most
people perceive some crucial difference. However, when it comes to enunciating just
what the difference is, we often make hand-waving references to how some exchanges
are “obnoxious” or some contracts “unconscionable.”

---

3
This paper seeks to spell out an economic principle of why certain kinds of
contracts, such as one where a worker is subjected to sexual harassment, may need to be
legally banned. I begin with some empirical context, discussing the magnitude of sexual
harassment in the workplace and the evolution of the law. The focus will be on U.S.
experience, since the United States has played a pioneering role in curbing workplace
harassment and the American law has been a model to many nations that have recently
drafted sexual harassments laws (like Bangladesh) or are in the process of drafting such a
law (like India). I then offer a theoretical model to explain why society may desire
legislative intervention to control sexual harassment in the workplace. It is important to
unearth the underlying principle for such rules, since it can influence a host of labor
market policies, such as stopping workers from being exposed to excessive health
hazards and having statutory limits on the hours of work. This economic approach is
then used to critique the current law and government policy, both with regard to sexual
harassment and in other matters of labor rights and standards.

The Context


In the United States, charges of sexual harassment are usually handled under Title
VII of the Civil Rights Act of 1964, which prohibits employment discrimination on the
basis of race, color, religion, sex or national origin. The Equal Employment Opportunity
Commission (EEOC), which went into effect in 1965, was established to enforce and
administer this statute.

---

4
The Incidence of Sexual Harassment
In 2001, 15,475 cases of sexual harassment were filed in the United States, shown
in Table 1. This total combines charges filed with the Equal Employment Opportunity
Commission and various state and local level Fair Employment Practices Agencies. Since
all cases that end up in the courts are filed with the EEOC or one of the state and local
agencies, this total offers a fair measure of the number of charges that are serious enough
to enter the legal process. The total number of sexual harassment charges rose sharply
from 1992 to 1995, and has remained at roughly the same level since then. This was
probably a lagged effect of the confirmation hearing of Clarence Thomas for the U.S.
Supreme Court, which brought high visibility to issues of sexual harassment in the
workplace, and also the passage of the Civil Rights Act of 1991. This Act, which became
effective in November 1991, allowed victims to claim compensatory damages for
suffering caused by discrimination. Prior to the 1991 act, the only forms of relief
available were “injunctive relief”—a court injunction that the discriminatory act be
stopped—or, in the most egregious cases, that the harassee, who may have been denied
promotion or dismissed, get back pay.

Table 1 also presents the number of cases “resolved” each year. (These figures
can exceed the number of charges filed in any particular year because of overhang of
charges from previous years.) A lack of resolution does not mean that the case did not
have merit since many cases are closed for administrative reasons, such as failure of the
charging party to respond to EEOC communications. Of the cases that are resolved, of
particular interest are the “merit resolutions,” since in these cases the allegations seemed
to have enough merit that the charges led to outcomes favorable to the charging parties.

---

5

The final row of the column shows cases where the harassment may be sexual in
nature but the basis of the harassment is not the plaintiff’s sex, but some other attribute,
such as the person’s race or religion. Hence, in 2001, 2.5 percent of sexual harassment
charges involved cases where the basis of harassment was not the plaintiff’s sex.
Of course, Table 1 should not be treated as a measure of the actual incidence of
sexual harassment in the workplace. The table does not include charges that do not reach
the EEOC and the state and local level Fair Employment Practices Agencies, because
they are filed and resolved within firms and corporations. Moreover, there is surely a
large number of cases where the victim does not file charges at all. On the other hand,
there must be some bloating of numbers caused by false charges.
A few sector-specific studies offer some sense of the percentage of people who
face sexual harassment in the workplace. The largest study was based on a survey of
active-duty women in the U.S. Armed Forces in 1995. A questionnaire was sent to 49,003
individuals and there were 28,296 returns, of which 22,372 were from women. According
to this study, 70.9 percent of active-duty women had faced some form of sexually
harassing behavior over the previous one year (Antecol and Cobb-Clark, 2002). This
figure seems very high, which may reflect the special circumstances of the armed forces
or be caused by the survey design.
One of the largest studies from the civilian sector in the United States was
conducted by the U.S. Merit Systems Protection Board in 1980, 1987 and 1994 among
federal employees all over the country (USMSPB, 1995). In 1994, a questionnaire was
sent out to nearly 13,200 federal employees and about 8,000 were returned. This study
found that 44 percent of the women employees and 19 percent of the male employees had

---

6
faced some form of sexual harassment in the previous two years. These numbers were
only slightly higher than in the previous surveys. In a similar survey in 1987, 42 percent
of women employees and 14 percent of male employees claimed to have faced some
form of sexual harassment; still further back in a 1980 survey, the figures were 42 percent
of women employees and 15 percent of male employees. The study also found that only
6 percent of those who faced harassment actually lodged a complaint (and, no doubt,
many fewer filed a legal charge).
Since sexual harassment is notoriously difficult to define, self-reported surveys
like these have to be treated with caution and more work remains to be done in measuring
the extent of sexual harassment (Welsh, 1999). For example, it is a plausible but
unproven hypothesis that the quantity of sexual harassment in the workplace has declined
in recent years and decades, but that an increasing awareness and empowerment of
women workers has led to a greater percentage of cases being reported in surveys. The
statistics we observe are a mixture of these countervailing forces. But despite the
uncertainties that surround estimates of the extent of sexual harassment, the figures are
sufficient to indicate that actual charges reported in Table 1 are the tip of the iceberg of
instances of sexual harassment in the workplace.

The Concept of Sexual Harassment in Law
One reason for the different definitions is that “sexual harassment” as a legal
concept is only about 25 years old. According to Farley (1978), the concept was
“discovered” in 1974 in the course of discussions in a class on women and work in
Cornell University (see also Crouch, 2001). We have come a long way since then. In the

---

7
years after the passage of the Civil Rights Act of 1964, federal courts typically refused to
view sexual harassment as a form of employment discrimination under the meaning of
the statute. However, in the case of Barnes v. Costle (561 F.2d 983 [D.C. Cir. 1977]), a
federal circuit court held that sexual remarks and solicitations, when linked to threats
about being fired, constituted sexual harassment. Catherine MacKinnon (1979), who was
one of the attorneys for Barnes, then published a pioneering book which argued that
sexual harassment itself, with or without a threat of being fired, constituted employment
discrimination on the basis of sex in the meaning of the Civil Rights Act of 1964.

In 1980, the Equal Employment Opportunity Commission issued “Guidelines on
Discrimination Because of Sex,” which declared sexual harassment a violation of Title
VII of the Civil Rights Act of 1964. Along with offering some guidelines for establishing
criteria for determining when sexual harassment has occurred, these guidelines split
sexual harassment into two categories: “quid pro quo” harassment, whereby a refusal to
grant sexual favors were met with blocked promotion or frozen wages or outright
dismissal from work; and “hostile environment” harassment, which took the form of
sexually abusive language or gestures which made some workers feel humiliated and
discriminated against.1
The courts soon followed the lead of the Equal Employment Opportunity
Commission. In Bundy v. Jackson in 1981 (641 F.2d 934, 942 n. 7 [D.C.Cir.1981]), a
hostile environment alone was for the first time recognized by a federal appeals court as a
form of harassment. In Bundy, a female employee of the Department of Corrections in
Washington, D.C., was repeatedly invited by her supervisor to describe her sexual
experience. When she complained about these comments to a senior manager, he took it

---

8
lightly, saying that the feelings of the supervisor were understandable. The court upheld
Bundy’s charge that the innuendo and implicit threats created an intimidating and hostile
atmosphere, and were unlawful, even though she had not suffered any tangible loss, such
as the withholding of salary increments or promotion.2
In 1986, the U.S. Supreme Court affirmed this distinction in Vinson v. Meritor
Bank case (477 U.S. 57 [1986]), in which Michelle Vinson, a trainee-teller, was
repeatedly propositioned by Sidney Taylor, a vice president of the bank. After resisting
for some time, she relented for fear of losing her job and was subjected to repeated
unwanted sexual relations for over four years. In this case the court did not find that the
worker had suffered in terms of pay or promotions; in fact, the court did not even find it
necessary to decide whether a sexual relationship between worker and manager had
happened at all. The court held that a hostile work environment alone was a violation of
employment discrimination under the Civil Rights Act of 1964.

A Model of Sexual Harassment in the Workplace

In building a model of labor markets for analyzing sexual harassment, I want to
motivate the exercise with a conceptual puzzle. One of the basic principles used by
economists to guide policy decisions is the “principle of free contract,” which asserts that

1 Juliano (1992) discusses the history of the “hostile environment” claim.
2 As a digression, note that the “hostile environment” clause suggests that what constitutes sexual
harassment may well have a cultural element to it and so may, reasonably, differ across time or across
nations. In the United States, defendants in harassment cases have sometimes tried to use the First
Amendment (Schauer, 2002), arguing that, for instance, Playboy posters in the workplace should be
allowed as a form of freedom of expression. Even if we were to contest this argument, most people would
recognize that there is a line where not all art or expression that is somehow related to sex or gender and
that offends a single person should be grounds for a sexual harassment suit. The point here is that different
societies may wish to draw the line in different places.

---

9
when two or more consenting adults agree to a contract or an exchange that has no
negative externalities on uninvolved individuals, then government has no reason to
intervene and prohibit such a transaction.3 This principle is the product of two more
fundamental ideas: the Pareto principle and consumer sovereignty. The Pareto principle
asserts that if a change is such that at least one party is better off and others are no worse
off, then that change is desirable. Consumer sovereignty asserts that each (adult)
individual is the arbiter of that individual’s own welfare.
Now consider a case where a firm, either by virtue of its reputation for harassment
or by writing down an explicit contract, ensures that a potential employee knows that she
will be sexually harassed on the job. If she nevertheless accepts the job, then, by the
principle of free contract, there seems to be no economic case for stopping such a
contract. From this it seems a short step to argue that government should not use the law
to stop sexual harassment in the workplace. It should be left to the individuals involved to
be worked into the terms of employment contract appropriately. Given the heterogeneity
among human beings, some will agree to work for lower wages but want a guarantee of
no harassment; others may prefer higher wages, while relinquishing the right not to be
harassed.4
However, I will argue that what was described as a “short step” in the above
paragraph is deductively invalid, because there is a difference between a single contract
or a small number of contracts of a particular kind and a large number of contracts of that

3 This principle has a long intellectual history and has been subjected to repeat scrutiny. John Stuart Mill
(1848) favored this principle, though was concerned about some special cases, such as a laborer’s right to
enter into voluntary slavery. For contemporary discussions in the context of labor markets, see Zimmerman
(1981), Trebilcock (1993), Sunstein (2001) and Kanbur (2002).
4 Observe that in this scheme each worker has the right not to be harassed, and in addition they have the
right to give up that right. As Sunstein (2001) lucidly demonstrates, giving a person the right to give up a
certain right, call it R, is not the same as not giving the right R (see also Basu, 1984).

---

Document Outline

• Sexual Harassment Charges Filed with Equal Employment Opportunity Commission and Fair Employment Practice Agencies, 1992-2001

---