. 23
( 32 .)


net effect is that organized labor tends to create a surplus of job
applicants. Given that surplus, the cost to the employer of turning
away qualified applicants from the "wrong" group is less than it
would be if he had to be concerned about finding enough similarly
qualified replacements for those who have been arbitrarily re-
Even in the absence of discrimination by guilds or unions them-
selves”and there was plenty of that”it would still be cheaper for
employers to discriminate on their own. Given this situation, it is

not so puzzling that Jewish artisans found it easier to practice their
skills in that part of Poland that was more hostile to Jews and that
black American artisans found it easier to practice their skills in
the Jim Crow South than in the more unionized North. Differ-
ences in costs of discrimination outweighed differences in hostile
The same pattern can be seen in employment statistics over
time. Both in the American South during the Jim Crow era and in
South Africa under white rule, blacks were a much higher percent-
age of railroad employees in the early twentieth century than they
were at mid-century. In both countries, labor markets were more
freely competitive in the earlier era and more controlled in the
later era”and in both countries it would be hard to claim that
there was less racism in the earlier era.
Not only labor unions, but also government regulation, can re-
duce the cost of discrimination. Where a public utility with a mo-
nopoly has its prices set by a government regulatory agency on the
basis of its costs, it has little or no incentive to keep those costs
down to a level that would be necessary for its survival in a com-
petitive market. Costs of discrimination, like other costs, can sim-
ply be passed on to the customers of a regulated monopoly. When
the American telephone industry was a regulated monopoly, few
blacks were hired for even such routine jobs as telephone operators
before the civil rights laws were enacted. Because each local tele-
phone company was a monopoly within its own territory, it could
pass on higher costs to everyone who used telephones. Had it not
discriminated, its costs would be lower and its monopoly profits
could theoretically have been higher, but because its profit rates
were in fact constrained by government regulation, the phone
company would never have seen that additional money anyway. In-
stead, it could indulge its racial preferences with no net loss of
The Economics of Discrimination 173

Meanwhile, blacks were beginning to star on Broadway as early
as the 1920s, in an industry with cut-throat competition, where
large profits and devastating losses were both common. The cost
of refusing to hire black entertainers who could fill a theater was
just too high for this industry to follow the same practices as the
telephone industry. The one-to-one correspondence between
racism and discrimination that is often assumed cannot explain
such differences between sectors of the same economy at the same
time. Even less can it explain the persistence of such differences
over time, when there is a complete turnover of decision-makers
throughout the economy. Even after a given set of decision-mak-
ers and their individual predispositions have passed from the
scene, the persistence of the same set of incentives tends to repro-
duce the same end results with a new set of decision-makers in the
same respective industries, whatever the individual predispositions
of these new decision-makers.
Because major league baseball operated as a cartel exempted
from anti-trust laws, it too had low costs of discrimination and
was able to keep black players out”so long as all teams did so.
But this situation changed in 1947, when the Brooklyn Dodgers
hired Jackie Robinson as the first black major league ballplayer.
Because there was competition within the cartel among its various
teams, once the color barrier was broken by just one team hiring
just one black player, the cost to other teams of keeping out other
black players rose sharply. The net result was that, in a matter of a
relatively few years, large numbers of black players flooded into
the major leagues. For a period of seven consecutive years, no
white man won the National League's Most Valuable Player
award. Had other teams not followed the lead of the Dodgers in
hiring black players, all these MVP stars would have become
Dodgers, giving Brooklyn a virtual monopoly of National League
pennants and perhaps of world championships.

This cost was obviously much too high for the competing teams
to pay for continuing racial exclusion in major league baseball.
Their racial attitudes may not have changed, but the cost of trans-
lating those attitudes into discriminatory exclusions had changed
Given the influence of the costs of discrimination on the amount
of actual discrimination, it is also possible to understand another
otherwise puzzling phenomenon”the especially strong reversals
of racial policies in sectors of the economy least subject to the pres-
sures of the competitive marketplace. These include the govern-
ment itself, government-regulated public utilities, and non-profit
organizations, such as academic institutions, hospitals, and foun-
dations. Colleges and universities that had never hired blacks for
their faculties in pre-World War II America led the way when af-
firmative action in the 1960s and 1970s meant preferential hiring
and promotion of black professors and preferential admissions of
black students. There was also a very similar sharp reversal of hir-
ing policies in the telephone industry, among others, at the same
These sudden radical changes from especially discriminatory
policies against a particular group to preferential policies toward
the very same group are hard to explain by predispositions, since
many of the same decision-makers were in control during the tran-
sition period. It is much easier to understand in terms of the incen-
tives and constraints of the circumstances in which they operated.
More specifically, neither discrimination nor "reverse discrimina-
tion" cost them as much as either policy would have cost decision-
makers in those sectors of the economy where institutional survival
depends on keeping costs within narrow limits, in order to meet
competition in a free market. Once the political and social climate
changed, government, government-regulated utilities, and non-
profit organizations could change most quickly with the least cost
to themselves.
The Economics of Discrimination 175

The power of the free market was perhaps best demonstrated in
white-ruled South Africa during the era of apartheid. Here we
need not wonder about racial predispositions or about the fact that
the vast majority of employers in industry, agriculture, and govern-
ment were white. Yet, even in a country which became a world-
wide symbol of racial oppression, white employers in competitive
industries violated official government policy on a massive scale by
hiring more black workers and in higher positions than the law al-
lowed. There is no compelling evidence that these particular white
employers had different racial predispositions than the white peo-
ple who administered the apartheid government. What they had
were very different costs of discrimination.
While government agencies and government-regulated rail-
roads, for example, could maintain apartheid policies at virtually
zero cost to themselves, it was a wholly different economic situa-
tion for people spending their own money. Home-building was a
typical example:

To build a house in Johannesburg meant either waiting for months
for a white, expensive, legal building gang, or finding a black gang,
perhaps with a white nominally in charge in case an official came
inquiring. Most customers opted for the quicker, cheaper service.

Such practices became so widespread in South Africa that the
white-run apartheid government cracked down in the 1970s, fin-
ing hundreds of building construction companies. Moreover, this
was by no means the only industry that hired more blacks than
they were allowed to by law. In the Transvaal clothing industry, no
blacks at all were allowed to work, under the apartheid laws. Yet,
as of 1969, blacks were an absolute majority of the workers in that
industry. There were also residential areas in South Africa set
aside by law for whites only”and yet there were not only many
non-whites living in these areas (including black American econo-

mist Walter Williams), at least one such area had an absolute ma-
jority of non-whites. Competition in a free market simply made
discrimination too expensive for many, even though violating the
apartheid laws also cost money.1
The expansion of black residential areas into white residential ar-
eas has been even more common in the United States. However, this
more or less continuous expansion of black ghettoes has been in
contrast to the history of the original ghettoes”those of the Jews in
Europe in centuries past. Jewish ghettoes in Europe in centuries past
tended to become more overcrowded as the Jewish population grew,
though there were particular times and places where Jews were al-
lowed to expand an existing ghetto or to set up a new ghetto to ac-
commodate their growing populations. Here again, the difference
has been in the economic costs of discrimination.
Black ghettoes have expanded through the marketplace because
of the costs of excluding black renters and home buyers. This is not
to say that there was no resistance by whites. Often there was or-
ganized, bitter, and even violent resistance. The key question, how-
ever, is: What was the end result? The usual end result was that
black ghettoes expanded in cities across the country. Moreover,
where this expansion was stopped by laws or government policies,
by restrictive covenants, or by violence or the threat of violence,
that reinforces the point that the costs of discrimination were too
great for the expansion of black ghettoes to be stopped in the mar-
ketplace. By and large, black ghettoes continued expanding with
the growth of the black population.

One of the reasons for the weakening of apartheid, even before the end of white-minority rule in
South Africa, was that many of the white Afrikaners, the principal supporters of apartheid, rose over
the years into the ranks of business owners and now had to pay the costs of discrimination, which
had been paid by British and Jewish business owners before. Faced with these costs, many Afrikan-
ers began to lose their enthusiasm for apartheid and some even spoke out against it, despite the
authoritarian South African government.
The Economics of Discrimination

The boundaries of Jewish ghettoes in Europe were not deter-
mined by the marketplace but were established by the dictates of
those with political power. Only when these political leaders
found it expedient did these boundaries expand. That is why Jew-
ish ghettoes tended simply to become more crowded with the pas-
sage of time and population growth. There was usually no cost to
political leaders for discriminating against Jews. In particular cir-
cumstances”when there was a war on, for example, and the rulers
needed the help of Jewish financiers”various proscriptions might
be eased and more ghettoes allowed to be founded to relieve over-
crowding. During the Thirty Years War (1618-1648), for exam-
ple, new Jewish communities were permitted to be established and
new occupations and markets opened up to Jews, while a syna-
gogue was permitted to be built in Vienna for the first time in
more than 200 years and a synagogue was permitted in Denmark
for the first time ever.
In short, costs of discrimination are not only a fact of life, they
are a potent force in actual decision-making, even in countries
with strong racial, ethnic, or religious predispositions. How much
of a force depends on the economic incentives and constraints in
particular sectors. What this means is that not only is the assumed
one-to-one correlation between racism and discrimination false,
but also that those who wish to reduce discrimination need to pay
attention to the economic conditions which make it more expen-
sive or less expensive for decision-makers to discriminate. Too of-
ten, those opposed to discrimination are also opposed to free
competitive markets that make discrimination more costly. They
do not think beyond stage one.
Even a given market”such as the market for housing, for ex-
ample”can have more discrimination or less discrimination ac-
cording to whether its prices are detemined by supply and demand
or are imposed by external agencies such as government, labor

unions, or a cartel. For example, when a landlord refuses to rent an
apartment to people from the "wrong" group, that can mean leav-
ing the apartment vacant longer. Clearly, that represents a loss of
rent”if this is a free market. However, if there is rent control, with
a surplus of applicants, then such discrimination may cost the
landlord nothing.
Similar principles apply in job markets. An employer who re-
fuses to hire qualified individuals from the "wrong" groups risks
leaving his jobs unfilled longer in a free market. This means that
he must either leave some work undone and some orders unfilled
or else pay overtime to existing employees to get it done, losing
money either way. However, in a market where wages are set artifi-
cially above the level that would exist through supply and demand,
the resulting surplus of applicants can mean that discrimination
costs the employer nothing. Whether these artificially higher
wages are set by a labor union or by a minimum wage law does not
change the principle.
In all these cases, the crucial factors in the cost of discrimina-
tion have been the presence or absence of competition and
whether those making the decisions have been spending their
own money or someone else's money. When one's own money is
at stake, groups hostile to each other may not only fail to discrim-
inate, they may in fact seek each other out, as Polish immigrants
and Jewish immigrants from Poland did in early twentieth-cen-
tury Chicago:

. .. the Poles and the Jews in Chicago .. . have a profound feeling of
disrespect and contempt for each other, bred by contiguity and by
historical friction in the pale; but they trade with each other on Mil-
waukee Avenue and on Maxwell Street. A study of numerous cases
shows that not only do many Jews open their businesses on Milwau-
kee Avenue and Division Street because they know that the Poles
The Economics of Discrimination

are the predominant population in these neighborhoods, but the
Poles come from all over the city to trade on Maxwell Street be-
cause they know that there they can find the familiar street-stands
owned by Jews.


Those who equate discrimination with differences in life chances


. 23
( 32 .)