. 25
( 32 .)


were classified. Even the term "affirmative action," as used initially
in President John F. Kennedy's Executive Order in 1961, meant
that an employer must be active, rather than passive, to make
sure that job applicants and employees were treated the same,
"without regard to their race, creed, color or national origin."
Anti-discrimination laws, such as the Civil Rights Act of 1964,
were likewise designed to prevent differential treatment in em-
ployment, in access to public accommodations, and in various
other contexts. However, even before the Civil Rights Act of 1964
was passed, a competing definition of discrimination had already
begun to emerge.
Earlier that year, a state Fair Employment Practices Commis-
sion in Illinois ruled that the Motorola company had discrimi-
nated against a black job applicant by requiring him to take the
same test that other applicants took. This was considered to be
discriminatory because blacks usually did not have either as much
or as good an education as whites, and so were less likely to do
well on tests. As one of the Commission's examiners put it, the
test was "unfair to culturally deprived and disadvantaged groups."
This implicitly shifted the operational definition of discrimination
from a denial of equal treatment to a denial of equal prospects of
Critics of the federal civil rights legislation being considered in
Congress pointed to the Motorola case as an example of what to
expect if the Civil Rights Act of 1964 became law, while the Act's
advocates and defenders vehementaly denied it. Although noth-
ing in the Act itself authorized any such definition of discrimina-
tion, subsequent interpretations by the federal courts, including
the Supreme Court, established the broader definition of dis-
crimination as the use of criteria which have a "disparate impact"
on the prospects of success by minority groups. In other words, a
failure to have a workforce reflecting the demographic profile of

the local community was now taken as presumptive evidence of
discrimination, with the burden of proof to the contrary falling on
the employer.
The issues revolving around such policies have already been am-
ply”or perhaps more than amply”debated by many advocates
and critics from political, moral, legal, and sociological perspec-
tives. The economic question, however, is: What incentives and
constraints does this definition of discrimination create and what
are the likely consequences of those incentives and constraints?

Economic Consequences

The most obvious and most immediate consequence is that those
employers who were discriminating against minorities now had an
incentive to stop doing so, in order to avoid the costs arising from
legal action under anti-discrimination laws. However, the empiri-
cal evidence suggests that the extent of discrimination in the sense
of treating equally capable individuals differently because of race,
may have been considerably less than that assumed by those who
equate discrimination with different outcomes for different
groups. Nevertheless, whether the amount of discrimination is
large or small, the immediate, stage-one effect of anti-discrimination
laws is to reduce it.
What about other consequences unfolding over time? The ease
with which private parties or government agencies can make a
charge of discrimination against an employer who is in fact not
discriminating, but whose work force does not reflect the local
population profile, now makes locating near concentrations of mi-
nority populations legally more risky. Those employers in these
risky locations may not be able to do much about it in the short
run. However, as time goes on, new businesses that arise have a
choice of where to locate and even old businesses with branches in
The Economics of Discrimination 189

more than one location can shift their operations away from mi-
nority population centers, in order to reduce their legal risks.
Eventually, even the employer with only one plant or office can
find an opportunity to move elsewhere.
Whether minority workers gain more jobs through the effect of
anti-discrimination laws in reducing discrimination or lose more
jobs through the incentives that such laws provide for employers
to relocate is an empirical question. But it is a question that is sel-
dom asked.
For the employer, one way of avoiding charges of discrimination
by minority employees or job applicants may be to hire in such a
way as to have a racially representative sample of the local popula-
tion. However, this can risk charges of "reverse discrimination" by
white job applicants who were not hired, or white employees who
were passed over for promotion. Since racial or ethnic groups tend
to differ in their particular skills and experience, it is unlikely that
the qualifications of members of different groups in a given loca-
tion will be the same”the only way to have both non-discrimina-
tory hiring and a demographically representative workforce at the
same time.
One way out of this dilemma for many employers is to hire in
such a way as to create demographic representation, and at the same
time escape charges of "reverse discrimination," is to justify prefer-
ential hiring on grounds of affirmative action based on a need for
"diversity." This legal avoidance of lawsuits from both minority and
majority individuals, or government agencies, works only in so far as
courts accept such arguments. Therefore, when affirmative action
has been challenged in the Supreme Court, even when it is chal-
lenged in cases involving college admissions, many corporations
have filed briefs in favor of the colleges with racial preferences in
admissions standards. The principle involved is a legal safe haven
for employers, and one which they do not want to lose.


Much discussion of discrimination fails to allow people with oppos-
ing views to confront their substantive differences”much less con-
front evidence”because so many of the words used have ambiguous
and shifting meanings. In particular, these discussions abound in
terms which confuse differences in internal personal qualities”such
as skills, education, and experience”with externally imposed barri-
ers to employment, college admissions, and other desired goals. Thus
those who fail to qualify for particular benefits are often said to be
denied "access" or "opportunity," when in fact they may have had as
much access or opportunity as anyone else, but simply did not have
the developed capabilities required.
The facts may of course vary from individual to individual or
from group to group, but we do not even know what facts to look
for when considering issues revolving around discrimination if our
terms make no such distinctions. Similarly, a mental test may be
characterized as "culturally biased" if one group scores higher than
another, as if it is impossible for different groups to have different
interest, experience, upbringing, education, or other factors that
would lead to a real difference being registered, rather than a mere
biased assessment being made.2
Would anyone consider basketball to be a culturally biased game
because blacks generally play it better than whites? It might well
be that, if whites took as much interest in the game and played it as
often as blacks, they would be just as good at it. But measuring
what people actually do in the real world is very different from as-

A complicating factor here is a history of controversies over whether there are innate racial differ-
ences in intelligence. But that explosive emotional issue has little or no relevance to questions about
discrimination. Innate potential exists at the moment of conception but no one applies for a job or for
college admissions at the moment of conception. By the time any such application is made, numerous
personal and social factors have had an influence over the years”and these factors have seldom been
the same for every individual or for every group. Moreover, employers are far more likely to be inter-
ested in the developed capabilities that have emerged, rather than the innate potential with which
someone began life.
The Economics of Discrimination 191

sessing what they would do under other, hypothetical circum-
stances. Nor is the owner of a professional basketball team engag-
ing in racial discrimination when he hires more blacks than whites
in a country where the population has more whites than blacks.
None of this says anything about how much discrimination there is.
What it says is that a vocabulary has come into being that makes it
virtually impossible to determine how much discrimination there
is, since the results of both discrimination and numerous other
factors are lumped together under the same words. Those who do
not think beyond stage one may assume that, whatever the level of
discrimination, anti-discrimination laws reduce it. Yet the incen-
tives and constraints created by such laws can either increase or re-
duce the employment or other opportunities of minority groups,
on net balance.
There may be honest differences of opinion on what the net bal-
ance turns out to be, but even the question itself does not arise un-
til we look beyond stage one. More precise definitions of terms or
finer breakdowns of statistics might yield very different estimates
of the amount of discrimination, but the crucial point here is that
the vocabulary actually in use is not adapted to discovering how
much discrimination exists, however well adapted it may be for
political purposes.
This page intentionally left blank
Chapter 7

The Economic Development
of Nations

Before 1886 there was not one cocoa tree in
British West Africa. By the 1930s there were
millions of acres under cocoa there, all owned
and operated by Africans.

”Peter Bauer

A ll over the world, national economies have generally ad-
vanced in recent centuries, producing more output per capita
and thereby raising people's standards of living. This advance in
technology and wealth has been greater in some regions of the
world than in others, and greater during some periods of history
than in others. However, there have also been periods of stagna-
tion and periods of retrogression, sometimes brought on by the
disruptions and destruction of war, sometimes by plagues or other
natural disasters, and sometimes by ill-conceived economic poli-
At the beginning of the twentieth century, Argentina was one of
the ten richest countries in the world, ahead of France and Ger-
many. But, at the beginning of the twenty-first century, no one
would even compare the troubled and chaotic Argentine economy


with that of France or Germany. Many ill-conceived policies in be-
tween led to this falling behind in a country blessed with natural
resources and spared the wars which ravaged other countries in the
twentieth century. Between 1998 and 2002, income per person in
Argentina dropped by two-thirds.1
Sharp changes in countries' relative positions have taken place in
much shorter periods than a century. As of 1991, India and China
had very similar output per capita. But, a decade later, China's out-
put per capita was double that of India. China had begun the
process of moving away from a government-run economy to more
of a market economy. When India began making the same kinds
of changes in the 1990s, its economic development became more
rapid as well.
While economic development has been more or less taken for
granted in much of the Western world since the industrial revolu-
tion of the eighteenth and nineteenth centuries, this has not been
so in all places and all times”and it was not so in the West for
many centuries before then. In some other parts of the world, even
in the twentieth century, farmers continued to farm and fishermen
continued to fish in much the same way as their ancestors had cen-
turies earlier, leading to standards of living not very different from
what they had been in ancient times.
In short, economic development has never been automatic. Why
it has been greater in some places than in others, and at some times
rather than others, is a question of great practical importance for the
economic fate of billions of human beings today, though it is a ques-
tion for which no single answer is likely to explain everything. Tech-
nology has had much to do with it, though not everything. Similarly
for geography, natural resources, and other factors, including the hu-

Anthony Faiola, "In Argentina, Dignity Turns to Despair," The Washington Post National Weekly
Edition, August 17-18,2002, p. 16.
The Economic Development of Nations 195

man factor. As in other areas of economics, attempts to understand
certain basic realities are made more difficult by the distractions of
popular myths and misconceptions.


The striking and even shocking differences in economic develop-
ment seen in many parts of the contemporary world have led
many to develop, or be drawn to, sweeping and often melodra-
matic theories that claim to explain such economic contrasts. Yet
such economic contrasts have been common throughout history.
Ancient China was so far in advance of Europe that, while there
was a great demand among Europeans for silk, chinaware (the
name is significant) and other exotic products, the Europeans pro-
duced nothing comparable to trade and so had to pay in gold for
what they bought from China. Within Europe, the contrasts have
been equally great. When the ancient Greeks had monumental ar-
chitecture that is still imitated today and landmark intellectual fig-
ures like Plato and Aristotle, illiteracy was common across much
of northern Europe and there was not a single building in all of
Britain when the Romans invaded in the second century A.D.,
nor had a single Briton's name yet entered the pages of history.
The contrast today between the wealth of Western countries and
the poverty of the Third World is nothing new.
While differences in economic development have long been
common, the particular advantages of one society over another
during a particular era have not been permanent. Europe eventu-
ally overtook China, after many centuries. So did Japan. More im-
portant, the advances achieved in one society have often diffused


. 25
( 32 .)