Some people are said to have discriminating tastes when they are especially discerning in detecting differences in qualities and choosing accordingly, whether choosing wines, paintings or other goods and services. But the word is also used in an almost opposite sense to refer to arbitrary differences in behavior toward people, based on their group identities, regardless of their actual qualities as individuals.
Both kinds of discrimination can result in large differences in outcomes, whether judging people or things. Wine connoisseurs can end up choosing one kind of wine far more often than another, and paying far more for a bottle of one kind of wine than for a bottle of the other.
Something similar can often be observed when it comes to people. It is common, in countries around the world, for some groups to have very different outcomes when they are judged by others in employment, educational and other contexts. Thus different groups may end up with very different incomes, occupations and unemployment rates, or very different rates of admissions to colleges and universities, among many other group disparities in outcomes.
The fundamental question is: Which kind of discrimination has led to such disparate outcomes? Have differences in qualities between individuals or groups been correctly discerned by others or have those others made their decisions based on personal aversions or arbitrary assumptions about members of particular groups? This is ultimately an empirical question, even if attempts to answer that question evoke passionate feelings and passionate certainty by observers reaching opposite conclusions.
Another way of saying the same thing is: Are group disparities in outcomes a result of internal differences in behavior and capabilities, accurately assessed by outsiders, or are those disparities due to external impositions based on the biased misjudgments or antagonisms of outsiders?
The answers to such questions are not necessarily the same for all group disparities, nor necessarily the same for a given group at different times and places. Seeking the answers to such questions is more than an academic exercise, when the ultimate purpose is to enable more fellow human beings to have better prospects of advancing themselves. But, before seeking answers, we need to be very clear about the words in which the question is asked.
At a minimum, we need to know what we ourselves mean when we use a word like “discrimination,” especially since it has conflicting meanings. The broader meaning—an ability to discern differences in the qualities of people and things, and choosing accordingly—can be called Discrimination I. The narrower, but more commonly used, meaning—treating people negatively, based on arbitrary aversions or animosities to individuals of a particular race or sex, for example—can be called Discrimination II, the kind of discrimination that has led to anti-discrimination laws and policies.
Ideally, Discrimination I, applied to people, would mean judging each person as an individual, regardless of what group that person is part of. But here, as in other contexts, the ideal is seldom found among human beings in the real world, even among people who espouse that ideal. If you are walking at night down a lonely street, and see up ahead a shadowy figure in an alley, do you judge that person as an individual or do you cross the street and pass on the other side? The shadowy figure in the alley could turn out to be a kindly neighbor, out walking his dog. But, when making such decisions, a mistake on your part could be costly, up to and including costing you your life.
In other contexts, you may in fact judge each person as an individual. But that this depends on context means that people have already been implicitly pre-sorted by the context, and only after that pre-sorting are they then judged as individuals. For example, a professor entering a classroom on the first day of the academic year may judge and treat each student as an individual. But that same professor, walking down a lonely street at night, may not judge and react to each stranger on the road ahead as an individual.
The students in a college classroom are not likely to be a random sample of the full range of variations found in the general population, and are more likely to represent a narrower range of people assembled there for a narrower range of purposes, and with a narrower range of individual characteristics, as well as being in a setting less dangerous than a dark street at night.
In short, one of the differences between the applicability of Discrimination I and Discrimination II is cost—and this is not always a small cost, nor a cost measured solely in money. Everyone might agree that Discrimination I is preferable, other things being equal. Nevertheless, one may still be aware that other things are not always equal, and sometimes other things are very far from being equal.
Where there is a difference in costs when choosing between Discrimination I and Discrimination II, much may depend on how high those costs are, and especially on who pays those costs. People who would never walk through a particular neighborhood at night, or perhaps not even in broad daylight, may nevertheless be indignant at banks that engage in “redlining”—that is, putting a whole neighborhood off-limits as a place to invest their depositors’ money. The observers’ own “redlining” in their choices of where to walk may never be seen by them as a different example of the same principle.
In short, Discrimination I can have prohibitive costs in some situations, especially when it is applied at the individual level. However, Discrimination II—the arbitrary or antipathy-based bias against a group, is not the only other option. Another way of making decisions is by weighing empirical evidence about groups as a whole or about the interactions of different groups with one another.
To take an extreme example, for the sake of illustration, if 30 percent of the people in Group X are alcoholics and 2 percent of the people in Group Y are alcoholics, an employer may well prefer to hire people from Group Y for work where an alcoholic would be not only ineffective but dangerous. This would mean that a majority of the people in Group X—70 percent—would be more likely to be denied employment, even though they are not alcoholics. What matters, crucially, to the employer is the cost of determining which individual is or is not an alcoholic, when job applicants all show up sober on the day when they are seeking employment.
This also matters to the customers who buy the employer’s products and to society as a whole. If alcoholics produce a higher proportion of products that turn out to be defective, that is a cost to customers, whether the customer gets the defective product or whether the number of products that have to be discarded at the factory because they are defective means higher prices for the products that are sold, because the costs of defective products that are screened out at the factory must be covered by the prices charged for the good quality products that emerge from that process and are sold.
To the extent that alcoholics are not only less competent but dangerous, the costs of those dangers are paid by either fellow employees or by the customers who buy dangerously defective products, or both. In short, there are serious costs inherent in the situation, so that either 70 percent of the people in Group X or employers or customers—or all three—are going to end up paying the cost of the alcoholism of 30 percent of the people in Group X.
This is certainly not judging each job applicant as an individual, so it is not Discrimination I in the purest sense. On the other hand, it is also not Discrimination II in the sense of decisions based on a personal bias or antipathy toward that whole group. The employer might well have personal friends from Group X, based on far more knowledge of those particular individuals than it is possible to get about job applicants, without prohibitive costs.
The point here is neither to justify nor condemn the employer but to classify different decision-making processes, so that their implications and consequences can be analyzed.
A real-life example of the effect of the cost of knowledge in this context is a study which showed that, despite the reluctance of many employers to hire young black males, because a significant proportion of them have had criminal records, those particular employers who automatically did criminal background checks on all their employees tended to hire more young black males than did other employers.1 In other words, where the nature of the work made criminal background checks worth the cost for all employees, it was no longer necessary to use group data to assess whether individual young black job applicants had a criminal background.
Discrimination I—basing decisions on empirical evidence—thus has two variations. The ideal, and more costly, variation is seeking and paying the cost for information that would permit judging each individual as an individual, regardless of the group from which that individual comes. This we can call Discrimination IA.
In other cases, where such information is too costly to be worth it, individuals may be judged by empirical evidence on the group they are part of. This can be called Discrimination IB. Both variations are different from Discrimination II, where the reason for treating individuals from different groups differently has no empirical basis but is due to personal bias or aversion to members of particular groups.
Even employers who have no animosity or aversions against particular groups may nevertheless engage in Discrimination IB—empirically based generalizations—when the employer knows that various groups react differently in the presence of other groups.
Back in nineteenth-century America, for example, when there were many immigrants from Europe in the workforce, some groups brought their mutual animosities in Europe with them to America. To have a workforce including both Irish Protestants and Irish Catholics working together was to risk animosities and even violence, with negative effects on productivity. In other words, a workforce consisting exclusively of either group was more efficient than a workforce consisting of both.
The same principle applies where different groups have especially positive reactions to one another. For example, the employer may be indifferent as to whether the work to be done is done by men or by women, and yet be well aware that men and women are not indifferent to each other, or else the human race would have become extinct long ago.
Therefore, in the interests of workforce efficiency, when a particular occupation is overwhelmingly chosen by women, such as nursing, the employer may be reluctant to hire a male nurse. Conversely, where lumberjacks are overwhelmingly male, the employer may be reluctant to hire a female lumberjack, even if she is demonstrably as fully qualified as the men.
Observers who point out that particular individuals are equally qualified, regardless of sex, miss the point. An equally qualified individual may do the work just as well as others, but if some of the others are distracted from their work, the net effect can be a less efficient workforce. That is the empirical basis that can lead employers to practice Discrimination IB, even if the employers have no bias or aversion to those less likely to be hired.
Misdiagnosing the basis for discrimination produces more than a difference in words. It can produce policies less likely to achieve their goals, or even policies that make matters worse. For example, forbidding employers from checking criminal records of job applicants can mean reducing the job opportunities of those young black males who have no criminal record.
Employment decisions are not the only decisions affected by discrimination of one sort or another. Where there are real differences between groups, with potentially dire consequences, such as murder rates several times higher in one group than in the other, Discrimination IB may be carried to the point of “redlining” a neighborhood or group, even when a majority of the group avoided are not guilty of the behavior feared.
Even in a high-crime neighborhood, for example, most people are not necessarily criminals.* But the costs of sorting the local population individually can be prohibitively high. Therefore decisions are likely to be made through a cruder decision-making process, relying on empirically based generalizations—Discrimination IB—rather than the more discerning, but costly, Discrimination IA or an antipathy-based or bias-based Discrimination II.
One of the consequences of such situations is that a law-abiding majority in a high-crime neighborhood can end up paying a high price for the presence of a criminal minority among compatriots living in their midst. Some businesses will not deliver their products—whether pizzas or furniture—to high-crime neighborhoods, rather than risk bodily harm, including death, to their drivers.
Taxi drivers may avoid taking passengers to such neighborhoods for the same reason, even when these are black taxi drivers refusing to go into black high-crime neighborhoods, especially at night. Supermarkets and other businesses are often reluctant to locate stores in such neighborhoods, for similar reasons.
All this hurts law-abiding people in high-crime neighborhoods, who are, in effect, paying a price for what other people are doing. In addition to being the principal victims of crimes committed by criminals in their midst, they also literally pay a price in hard cash for the behavior of others, in the higher prices usually charged for goods sold in neighborhoods where there are higher costs of doing business, due to higher levels of shoplifting, vandalism, burglary, pilferage and robbery—and higher business insurance premiums because of these and other neighborhood disorders.2
A study titled The Poor Pay More saw the poor in general as “exploited consumers,”3 taken advantage of by stores located in low-income neighborhoods. This view was echoed in the media, in government and in academic publications.4 Yet, because many low-income neighborhoods are also high-crime neighborhoods, The Poor Pay More committed an all too common error in assuming that the cause of some undesirable outcome can be determined by where the statistical data were collected.
In this case, researchers collected price data in the neighborhood stores. But the causes of those high prices were not the people who posted those prices in the stores. Moreover, while prices were higher in inner-city, low-income neighborhood stores, rates of profit on investments in such stores were not higher than average but lower than average.5
For people unaware of these facts, the higher prices may be seen as simply “price gouging” by “greedy” store owners—Discrimination II against minority neighborhoods. For those who see the situation this way, higher prices may appear to be a problem that the government could solve by imposing price controls, as a Harlem newspaper suggested during the 1960s furor over revelations that “the poor pay more.”6 But if businesses in these neighborhoods do not recover their higher costs of doing business there in the prices they charge, they face the prospect of having to go out of business.
There is often a dearth of businesses in low-income, high-crime neighborhoods, which would hardly be the case if there were higher rates of profit being made from the higher prices charged in such neighborhoods.
It may be no consolation to those law-abiding citizens in a high-crime neighborhood that the higher prices they have to pay are reimbursing higher costs of doing business where they live. Meanwhile, politicians and local activists have every incentive to claim that the higher prices are due to discrimination, in the sense of Discrimination II, even when in fact the community is simply paying additional costs generated by some residents in that community.
Those local residents who created none of those costs can be victims of those who did, rather than being victims of those who charged the resulting higher prices. This is not just an abstract philosophical point or a matter of semantics. The difference between understanding the source of the higher prices and mistakenly blaming those who charged those prices is the difference between doing things to lessen the problem and doing things likely to make the problem worse by driving more much-needed businesses out of the neighborhood. The difference between Discrimination IB and Discrimination II is not just an academic distinction.
Although higher prices in low-income neighborhoods are often discussed in the context of racial or ethnic minorities, the same economic consequences have been found where the people in the low-income neighborhoods were white. As the Cincinnati Enquirer reported: “Residents of eastern Kentucky refer to the higher prices and interest rates common in their area as the ‘hillbilly tax.’”7
Among the things that might be done to reduce the burden of unfairness to law-abiding residents of high-crime neighborhoods could be stronger law enforcement by the police and the courts. But, to the extent that the public—both inside and outside the affected communities—sees the high prices as Discrimination II against the affected community as a whole, due to bias or antipathy by the larger society, the imposition of stronger law enforcement may be seen as just another imposition of injustice on the affected communities.
In short, whether people believe that higher prices in low-income, high-crime neighborhoods are due to Discrimination II or to empirically-based decisions matters in terms of which policies to reduce the unfair burdens on law-abiding residents are politically feasible. Community or ethnic solidarity can be a major obstacle to seeing, believing or responding to the facts.
None of this tells us how much Discrimination I or Discrimination II exists in a given society—or how many disparities in outcomes are due to some other circumstances or some other decision-making process.
In some situations, there may clearly be costs deliberately imposed on a group by outsiders—Discrimination II—such as denying black American citizens the right to vote in many Southern states in times past. The racial segregation laws in those states, forcing black passengers to sit in the back of buses and trolleys, and denying them admissions to state universities, were obvious examples of a wider range of clearly racial discrimination, in the sense of Discrimination II.
The original European ghettos in centuries past, which forced Jews to live in a confined area and banned them from most European universities, were other examples of the same Discrimination II. Innumerable other groups in countries around the world—the “untouchables” in India being a classic example—faced even more and worse restrictions and oppressions.
These are all costs imposed by Discrimination II, and paid by its victims. What also warrants analysis, in order to understand cause and effect, are the costs paid by the discriminators, because these costs are factors in how much Discrimination II can persist in particular circumstances and institutions. Such costs have no such moral, political or ideological attraction as the costs paid by victims. But the costs that discriminators have to pay—and the various circumstances in which they do or do not have to pay those costs—can affect how much Discrimination II is in fact likely to be inflicted.
Understanding the costs paid by discriminators also presents opportunities for policies that can ensure that these costs cannot be evaded, as well as warnings that other policies may inadvertently free discriminators from paying such costs, if the circumstances are not understood.
Neither the amount nor the severity of Discrimination II is fixed permanently. It varies greatly from country to country and from one era to another in the same country. There was an era in which many American employers’ advertisements for some jobs said “No Irish Need Apply” or “Whites Only.” There was a time when some shops in Harlem, back when that was an upscale white community, had signs that read, “No Jews, and No Dogs.”8
Nor were Americans unique. In many other places and times around the world, group discrimination—that is, Discrimination II—was so pervasive and so widely understood that no such signs were necessary. For a woman, a Jew or members of some other groups to apply for certain jobs would have been considered a presumptuous waste of the employer’s time.
Discrimination II in hiring and promotion raises questions about both causation and morality. Both kinds of questions deserve to be examined—separately.
In trying to understand the causes and the consequences of discrimination in hiring and promotion, it is necessary to again consider whether this is Discrimination I or Discrimination II. This is not always an easy question to answer, and in fact easy answers such as automatically equating statistical disparities in outcomes with Discrimination II can be a major obstacle to getting at the truth.
An employer who judges each job applicant individually, without regard to the applicant’s group membership, can nevertheless end up with employees whose demographic makeup is very different from the demographic makeup of the local population.
One major demographic fact that is often overlooked by those who automatically equate statistical disparities in outcomes with Discrimination II is that different ethnic groups have very different median ages. As already noted, Japanese Americans have a median age more than two decades older than the median age of Mexican Americans.9 Even if every individual of the same age had the same income, regardless of which group that individual was part of, nevertheless there would still be serious disparities in income between Japanese Americans and Mexican Americans—as well as between many other groups.
A group with a median age in their twenties will obviously not have nearly as large a proportion of their population with 20 years of work experience as a group whose median age is in their forties. One group may therefore have a disproportionate number of people in high level occupations requiring long years of experience, while the other group may be similarly over-represented in professional sports or in violent crimes, both of which are activities disproportionately engaged in by the young.
Such disparities in outcomes are not automatically evidence of either outsiders’ biases or internal deficiencies in the groups. Either or both may be present or absent, but that requires empirical evidence going beyond gross statistical differences in outcomes.
In short, conditions prior to job applicants reaching an employer can have a “disparate impact” on the chances of someone from a particular group being hired or promoted, even if the employer judges each applicant on that applicant’s own individual qualifications, without regard to the group from which the applicant came. Put differently, even Discrimination IA can produce “disparate impact” statistics, just as Discrimination II can.
Age is just one of those pre-existing conditions. As already noted, children raised in families where the parents have professional occupations hear nearly twice as many words per hour as children raised in working-class families, and more than three times as many words per hour as children raised in families on welfare.10
Can we believe that such differences—and others—compounded over many years while growing up, make no difference in individual capabilities and social outcomes, when those children become adults seeking employment? All these individuals may have been very similar at birth, but many things happen between birth and applying for a job or for college admissions. And those things seldom happen the same for everybody. As we have seen, they happen differently for children born and raised in the same family.
More generally, you cannot tell where the disparity in outcomes originated from where the statistics were collected. The data may be collected at the employer’s place of business, but that does not preclude disparities in individual capabilities from having originated in the home in or in the local cultures where people were raised.
Not only differences in child-rearing, but also decisions made by individuals themselves, affect their outcomes. When more than three-quarters of all college degrees in education go to women and more than three-quarters of all college degrees in engineering go to men,11 the statistical predominance of women in teaching and men in engineering cannot automatically be attributed to employers’ biases. More fundamentally, the cause of a given outcome is an empirical question, whose answer requires untangling many complex factors, rather than simply pointing dramatically to statistical disparities in outcomes.
It is easy to understand how being denied an opportunity to be hired or promoted for some jobs can lead to some groups having lower incomes than others, and why that can arouse moral objections, not only from those denied jobs but also from others who find such practices morally repugnant.
From a causal perspective, other questions arise as to the reasons for such practices. Here the cost of discrimination to the discriminator plays a causal role in the outcome. There is also a cost for society at large. A society in which women are arbitrarily banned from many kinds of work can pay a huge cost by forfeiting the productive potential of half its population.
“Society,” however, is seldom a decision-making unit, except perhaps at election time or during a mass uprising. To understand decisions in general, or employment decisions in particular, requires understanding the incentives and constraints confronting the particular decision-makers in particular kinds of institutions, who cannot simply choose to do whatever they wish, without regard to the costs of their decisions to themselves.
In a competitive market for labor, or for the sale of the employers’ products, the validity of the beliefs behind a business owner’s decisions can determine whether that business operates at a profit or a loss, or whether it survives or is forced to go out of business. In short, we cannot simply go directly from attitudes to outcomes—even if these attitudes involve racism or sexism—as if there were no intermediate factor of costs for decisions made in a competitive market. A systemic analysis of markets cannot proceed as if there were no other factors involved besides what individual decision-makers happen to prefer.
Economists who have recognized this have ranged from followers of Adam Smith to followers of Karl Marx. The point was perhaps best expressed by Friedrich Engels, co-author with Marx of The Communist Manifesto. Engels said: “what each individual wills is obstructed by everyone else, and what emerges is something that no one willed.”12 An analysis of systemic causation is concerned with what emerges.
Adam Smith, patron saint of free-market capitalism, likewise had a systemic analysis of causation. He did not attribute the benefits of a capitalist economy to good intentions by capitalists.13 On the contrary, a case could be made that Adam Smith’s view of capitalists as individuals was even more negative than that of Karl Marx.14 Smith and Marx reached opposite conclusions as to the benefits or harm done by free-market capitalism, but neither based his conclusions on the intentions of capitalists. Each based his conclusions on the systemic incentives and constraints of economic competition.
Too many other observers, including some academic scholars, reason as if intentions automatically translate directly into outcomes. Thus, in his book The Declining Signif icance of Race, sociologist William Julius Wilson pointed out the various organized ways in which white Southern landowners and employers in the post-Civil-War South sought to keep down the earnings of black workers and black sharecroppers.15 But there was no reference in that book to empirical evidence on how those intentions actually turned out—in other words, on “what emerges,” as Engels put it.
By contrast, economist Robert Higgs, who researched the actual consequences of those efforts of white employers and landowners in the postbellum South, found that such organized efforts often collapsed, as a result of competition among white employers and landowners for black workers and sharecroppers.16 It might seem as if newly freed blacks—desperately poor, usually illiterate and unfamiliar with working as free people in labor markets—would be easy prey for whites united to enforce whatever wage and sharecropper conditions they wanted. But that ignores the inherent, systemic competitive pressures in a market economy.
In agriculture, especially—and the South was largely agricultural at the time—there is an inherent urgency about getting the land plowed and the seeds planted in the spring, or else there will be no crop in the fall. Those white landowners who were the first to violate the terms to which other white landowners sought to limit black workers and sharecroppers stood to be the first to assure themselves of a workforce sufficient in both quantity and quality to maximize the size of the crop that could be grown on a given piece of land.
Other white landowners, who stuck by the restrictions and/or who cheated the black workers and sharecroppers in various ways, tended to find themselves having to make do with whatever quantity and quality of black workers and sharecroppers remained, after other white landowners had skimmed the cream by paying higher wages and higher crop shares, in order to improve their own prospects of a profitable crop. Black workers and sharecroppers did not have to know economics in order to know where friends and relatives were getting a better deal, and go there.17
It is hardly surprising that these efforts at suppressing black workers’ pay and black sharecroppers’ shares often broke down under such economic pressures. “What emerges” in this case was that black incomes per capita in 1900 were, at a minimum, “almost half again” higher than they had been in 1867–68. This represented a rate of growth higher than that in the American economy as a whole during that period.18 Because they started from a far lower economic level, blacks remained poorer than whites. But Professor Higgs’ data indicated that “black incomes grew more rapidly than white incomes over the last third of the nineteenth century.”19
Businesses in general, making decisions in a labor market or a product market, are not like professors voting at a faculty meeting, because those votes seldom have any costs for the professors themselves, despite whatever good or bad results such votes may have for students or for the academic institution. The difference is the difference between decisions made subject to consequential feedback in a competitive market and decisions made with insulation from such feedback in academia and in other insulated venues.
To avoid endless and inconclusive debates about the presence or magnitude of racism, we can test our hypotheses about the costs of discrimination in a historical context where there was no ambiguity on the subject—namely, South Africa during the era of apartheid, ruled by a white minority government, elected with the black majority denied a vote, and openly promoting white supremacy.
Apartheid laws limited how many blacks could be employed in particular industries and occupations, and forbad their being hired for work above certain levels in those industries and occupations. Yet white South African employers in competitive industries often hired more blacks than they were allowed to under the apartheid laws, and in higher occupations than those laws permitted.
A government crackdown during the 1970s led to hundreds of firms in South Africa’s construction industry being fined for violating those laws. Nor was the construction industry the only one in which competitive businesses were fined for hiring more blacks, and in higher occupations, than allowed under the law. In some other industries blacks even outnumbered whites in particular job categories where it was illegal for blacks to be hired at all.20
There is no compelling evidence that the white employers violating those laws had different racial views than the white legislators who passed such laws. What was different was that employers who failed to hire black workers whom it was profitable to hire paid a price for Discrimination II, in the form of lost opportunities to make money, while the legislators who passed laws imposing Discrimination II paid no price at all. Indeed, legislators who failed to pass such laws could pay a price politically, in a situation where only whites could vote, and white workers wanted protection from the competition of black workers.
Both the employers and the legislators were rationally pursuing their own self-interests. It was just that the institutional incentives and constraints were different in a competitive market from the incentives and constraints in a political institution. Nor were labor markets the only markets affected by the costs confronting discriminators in South Africa.
Apartheid laws also made it illegal for non-whites to live in certain areas set aside by law for whites only. Yet many non-whites in fact lived in these whites-only areas. These included black American economist Walter E. Williams during a three month stay in South Africa, doing research.21 There was at least one whites-only area in South Africa where non-whites were a majority of the residents.22 Here again, costs were the key. The costs to owners of rental property in whites-only areas who forfeited economic benefits available by renting to non-whites competed with the costs of disobeying apartheid laws, and the latter did not always prevail.
While racists, by definition, prefer their own race to other races, individual racists—like other people—tend to prefer themselves most of all. That is what led to widespread violations of apartheid laws by white employers and landlords in competitive industries in South Africa. It cost nothing for white South Africans to vote for candidates promoting white supremacy. But the costs of refusing to hire black workers who would make their business profitable could be considerable.
Moreover, the cost of refusing to hire them when their competitors in the product market were hiring them, and therefore could have lower costs of doing business—enabling competitors to undercut the product prices of employers who failed to hire black workers—exposed those who obeyed apartheid laws to the risk of losing profits and potentially losing their whole business.
This is not to say that discriminatory laws and policies have no effect. There are costs to disobeying laws, as well as countervailing costs to following such laws, so outcomes depend on particular circumstances in particular times and places.
However, the costs of Discrimination II to the discriminator—in the context of a competitive market—are far lower, or even nonexistent, in situations where free market competition does not exist, such as in (1) public utility monopolies whose prices and profit rates are directly controlled by government, (2) non-profit organizations, and (3) government employment. In all these particular situations, Discrimination II has tended to be far more common than in competitive markets, not only in South Africa under apartheid, but also in other countries around the world.23
Unless one believes that decision-makers in these particular institutions have different racial or other views than decision-makers in competitive markets, and that such differences persist over time, as new generations of decision-makers come and go, the reasons for such institutional differences must be sought in particular incentives and constraints growing out of differences in the circumstances of those institutions.
One of the landmark struggles in the civil rights movement in mid-twentieth-century America was a campaign against laws in most Southern states mandating that black passengers sit or stand only in the back of buses, with the seats up front being reserved for whites. Although many people on both sides of this struggle regarded these laws as though they had existed from time immemorial, they had not. The history of such laws illustrates again the different roles of economic incentives and constraints versus political incentives and constraints.
Three decades after the end of slavery, laws mandating racially segregated seating in municipal transit vehicles began to be passed in many Southern communities, toward the end of the nineteenth century. The political situation had changed from that in the period immediately after the Civil War, when U.S. troops were stationed in Southern states, and Southern governments were subject to federal policies granting blacks the right to vote during what was called the Reconstruction era.
With the end of Reconstruction, and the return of local self-government in the South, blacks often lost the right to vote, by methods ranging from laws to organized terrorism. Racially segregated seating on municipal transit vehicles was just one of the political consequences. Before these laws were passed, it was common for blacks and whites to sit wherever they felt like sitting on public transportation vehicles.
Many, if not most, of the streetcar companies during that era were privately owned, and their profits depended on how many people—whether black or white—chose to ride in their vehicles, and how often.
The decision-makers in these privately owned companies understood that they could lose profits if offending black customers by making them sit in the back, or to stand when all the back seats were taken, even if there were vacant seats in the front section that was reserved for whites. Indeed, racially segregated seating could even offend some whites, when all the white section seats were filled but there were vacant seats in the section set aside for blacks.
In short, racially segregated seating in municipal transit vehicles was seen, by those who owned or managed such companies, as something that reduced profits. Not surprisingly, municipal transit companies in the South fought against the passage of laws requiring racially segregated seating in streetcars. After losing politically in the legislatures, some municipal transit companies then took the issue into the courts, where they lost again. Then, after the laws went into effect legally, many Southern municipal transit companies simply did nothing to enforce racially segregated seating. In some places, passengers continued for years to sit wherever they felt like sitting.24
Eventually, however, Southern government authorities cracked down. They began charging municipal transit company employees with violations of the law for not enforcing racially segregated seating, and in some cases the owners of those companies were threatened with prosecution if those transit lines did not enforce the racial segregation laws. Only then did the laws that had been passed, in some cases years earlier, finally get enforced.25
Railroads were also affected economically by racial segregation laws. When black and white passengers had to be carried in separate coaches, this imposed the considerable cost of buying additional coaches for their passenger trains, as well as the further additional costs of more fuel to move the now heavier trains.
This was especially costly where there were insufficient passengers to fill one coach. If there were only enough black and white passengers for the total to fill two-thirds of the seats in a coach, racial segregation laws could create a situation where there were now two coaches required, each with only one-third of the seats occupied.
Like municipal transit companies, railroad managements in the South opposed the racial segregation laws, in their own self-interest, even if their racial views might have been no different from those of politicians who passed such laws. But the incentives to which the politicians responded were votes—that is, white votes—while the incentives to which railroad owners and managers responded were financial, and money was the same, regardless of the racial source.
The famous Supreme Court case of Plessy v. Ferguson in 1896 arose from the cooperation of the railroads with Homer Plessy, who was challenging these racial segregation laws, in order to create a test case.
Although Plessy was part of the black community, he was genetically far more Caucasian than African, and was physically indistinguishable from white men. Had he simply gotten on a train and ridden to his destination, there was little likelihood that he would have been questioned about being seated in a railroad car set aside for whites only. But the attorneys for the railroad and the attorneys for Plessy cooperated in arranging a legal confrontation, so that there would be a case to take into the federal courts.26 Unfortunately for both, and for the cause of equal rights in general, the Supreme Court majority ruled against them.
There is no predestined outcome of the conflict between economic and political forces. What is important is to recognize the implications of that conflict when crafting or changing laws and policies.
It is not only in political institutions, but also in some economic institutions, that decision-makers are insulated from having to pay the costs of Discrimination II that they impose on others. Public utility monopolies, whose prices and profit rates are directly controlled by government regulatory agencies, are among the institutions insulated from paying the economic costs which a competitive market imposes on discriminatory behavior, whether directed against ethnic minorities, women or others.
Although engaging in Discrimination II when hiring employees could mean lower profits for a firm operating in competitive markets, a government-regulated public utility that has a monopoly in its field would not be allowed to earn a higher profit rate than the government agency deemed proper in any case. So a public utility was not forfeiting any additional profit that it would be allowed to keep, if it hired without regard to the group from which job applicants came.
Discrimination II might require the regulated public utility company to pay additional labor costs from having to offer higher salaries, in order to attract a larger pool of qualified applicants, from which only applicants from groups that the decision-makers preferred would be hired. But, as a government-regulated monopoly, such costs could be passed on to customers who had little choice but to pay those costs.
The history of the telephone industry, back when telephones meant land lines, and all the major phone companies in the United States were subsidiaries of the American Telephone and Telegraph Company (A.T.&T.), illustrates this pattern.
As of 1930, there were only 331 black women in the entire country working as telephone operators, out of more than 230,000 women in that occupation. As late as 1950, black women were still only one percent of all women working for phone companies.27 However, after creation of “fair employment practices” laws in some Northern states in the 1950s, and then federal civil rights laws and policies in the 1960s, many telephone companies reversed their policies, and blacks began to be hired disproportionately.
Prior to the 1960s, however, the state “fair employment practices” laws existed solely outside the South. Although a national sample of employment in the telephone industry showed that the employment of black telephone operators increased more than three-fold between 1950 and 1960,28 it was 1964 before the first black telephone operator was hired by phone companies in such Southern places as New Orleans, Florida or South Carolina.29
These regional differences reflected the fact that individual telephone companies were regulated by state governments, and reflected state political forces. In the South during the 1950s, in all 11 states that had once formed the Confederate States of America, the share of blacks among male employees of telecommunications companies actually declined during the same decade when telephone companies in Northeastern and Midwestern states were hiring more blacks.30
Later data from a national sample of telephone companies showed that blacks accounted for one-third of the total growth in telephone companies’ employees from 1966 to 1968—a trend that had begun in the 1950s and was concentrated mainly in Northeastern and Midwestern companies.31 Since all major telephone companies in the country were owned and controlled by A.T.&T., such sharp regional disparities in individual phone company policies were far more consistent with regional political differences between Southern and non-Southern state governments that regulated these companies, rather than with policies handed down from A.T.&T.’s national management.
What was consistent in all these various regions was that additional costs entailed by either preferential or discriminatory treatment of black job applicants were costs that phone companies could pass on to customers, who had little choice but to pay them, since there were only land lines at the time, and each phone company was a monopoly in its own area.
It was much the same story in the government-regulated oil and gas public utilities at that time, where that regulation was also by state agencies, and increased hiring of blacks was confined to states outside the South during that era.32 They too paid no cost for discriminating against blacks before, nor any cost for preferential hiring of blacks afterwards. The same was true of decision-makers who ran non-profit organizations or officials in charge of government hiring policies.
Similar incentives produced similar outcomes in nonprofit organizations such as academic institutions, hospitals and foundations—and different outcomes in profit-based businesses operating in competitive markets. Like decision-makers in regulated public utilities, those in non-profit organizations were able to go along with whatever the prevailing opinions and pressures of the time might be, without having to worry about the costs created by Discrimination II against minorities, which their institutions would have to pay.
Against this background, it is hardly surprising that employment discrimination against blacks and Jews was especially widespread among colleges, universities, hospitals and foundations until after World War II, when a revulsion against Nazi racism set in. Before that happened, however, there were 300 black research chemists employed in private businesses in this earlier era, but only three black Ph.D.s in any field employed by white universities.33
As for Jews, they were seldom found on American college and university faculties before World War II. Although Milton Friedman had a temporary academic appointment before the war, it lasted only one year, despite high praise for his work by students and colleagues alike, and he spent the war years working as a statistician before eventually becoming a regular, tenured professor of economics at the University of Chicago after the war.34
At about the same time, the University of Chicago had its first black tenured professor.35 The University of Chicago was exceptional only in doing such things before most of the rest of the academic world.*
Decades later, after the political climate had changed considerably, colleges and universities engaged in preferential hiring of black faculty, as well as preferential admissions of black students—again, without the academic decision-makers paying any price for their decisions, just as they paid no price for opposite policies earlier. “Affirmative action” in academia was sooner and more sweepingly adopted than in private industries operating in competitive markets.
All this happened too fast for such sweeping policy reversals in non-profit organizations to have been due to changing personnel in the role of decision-makers. In many, if not most, cases the same decision-makers who had discriminated against blacks were now instituting preferential policies favoring blacks. In neither case was the policy necessarily due to the personal beliefs, biases or values of the individual decision-makers, nor was the change necessarily due to “road to Damascus” conversions of personal views occurring among innumerable decision-makers at the same time.
In addition to laws and policies directly concerned with Discrimination II, other laws and policies with very different purposes can also change the amount and impact of adverse consequences on groups defined by race, sex or other characteristics. In short, unintended consequences can affect outcomes as readily as intended consequences, and sometimes even more so. Minimum wage laws and building restrictions are two examples among others.
Although minimum wage laws in the United States apply without regard to race, that does not mean that their impact is the same on blacks and whites alike. Where rates of pay are determined, not by supply and demand in a free market, but are imposed by minimum wage laws, that can affect the cost of Discrimination II to the discriminator.
A wage rate set above where it would be set by supply and demand in a freely competitive market tends to have at least two consequences: (1) an increase in the number of job applicants, due to the higher wage rate, and (2) a decrease in the number of workers actually hired, due to labor having been made more expensive. In this situation, the resulting chronic surplus of job applicants beyond the number of jobs available reduces the cost of refusing to hire qualified job applicants from particular groups, so long as the number of qualified job applicants refused employment is not greater than the number of surplus qualified applicants.
When, for example, the number of qualified black job applicants refused employment can be easily replaced by otherwise surplus qualified white or other job applicants, that reduces the cost of Discrimination II to the discriminating employer to virtually zero. On the most basic economic principles, such a situation makes racial or other discrimination far more affordable by employers, and therefore more sustainable, than in a situation where wage rates are determined by supply and demand in a free, competitive market.
In the latter case, where supply and demand leave no chronic surplus or chronic shortage of labor, qualified black job applicants turned away have to be replaced by attracting additional other qualified job applicants from other groups by offering higher pay than what that pay would be by supply and demand in a freely competitive and non-discriminatory labor market. In other words, Discrimination II has costs to discriminators in a free market, greater than its costs when a minimum wage law creates a chronic surplus of job applicants.
Empirical evidence is consistent with this hypothesis. The prevailing national minimum wage law in the United States is the Fair Labor Standards Act of 1938. However, high rates of inflation that began in the 1940s put virtually all money wages above the level specified in that Act. As a practical matter, there was no minimum wage in effect a decade after the law was passed. As economist George J. Stigler pointed out in 1946, “The minimum wage provisions of the Fair Labor Standards act of 1938 have been repealed by inflation.”36
As of 1948, during this period of no effective minimum wage law, the unemployment rates of both black and white teenagers were just a fraction of what they would become in later years, as minimum wage rates began rising in the 1950s, in order to try to catch up, and then keep up, with inflation in later years.
What is particularly striking, however, is that there was no significant difference between the unemployment rates of black and white teenagers in 1948. The unemployment rate for black 16-year-old and 17-year-old males was 9.4 percent. For their white counterparts, the unemployment rate was 10.2 percent. For 18-year-old males and 19-year-old males, the unemployment rate was 9.4 percent for whites and 10.5 percent for blacks. In short, there was no significant racial difference in unemployment rates for teenage males in 1948,37 when there was no effective minimum wage.
After the effectiveness of the minimum wage law was restored by recurring minimum wage increases in later years, not only did teenage unemployment rates as a whole rise to multiples of what they had been in 1948, black teenage male unemployment rates became much higher than the unemployment rates for white teenage males—usually at least twice as high for most years from 1967 on into the twenty-first century.38
Labor force participation rates tell much the same story. As of 1955, labor force participation rates were virtually the same for black and white males, aged 16 and 17. For 18-year-old and 19-year-old males, blacks had a slightly higher labor force participation rate than whites, as was also true of males aged 20 to 24. But this pattern changed drastically, as minimum wage rates rose over the years.
In the mid-1950s, black labor force participation rates for 16-year-old and 17-year-old males began falling below that of their white counterparts, and the gap grew wider in succeeding decades. For males aged 18 and 19, the same racial reversal in labor force participation rates occurred a decade later, in the mid-1960s. For males aged 20 to 24, that same racial reversal occurred at the beginning of the next decade, in 1970. The magnitude of the racial difference in labor force participation rates among males, after the racial reversal, followed the same pattern, being greatest for the 16-year-olds and 17-year-olds, less for males aged 18 and 19, and least for males aged 20 to 24.39
These labor force participation patterns shed additional light on the basis for racial differences in employment. If the primary reason for that racial difference in labor force participation rates was racism (Discrimination II), there was no reason for such reversals, and especially reversals in different years and with different magnitudes for different age groups. People who are black at age 16 remain black as they get older, so there is no basis for racists to change their treatment of blacks in such patterns as black workers age.
If, however, the real reason for these patterns was that the work experience and job skills of younger black workers made them less in demand than older black workers with more work experience and/or more job skills, then a rising minimum wage rate prices the younger blacks out of jobs first and to the greatest extent.
Unfortunately, when minimum wage laws reduce the employment prospects of inexperienced and unskilled black teenagers, that reduces their labor force participation, and therefore reduces their rate of acquisition of work experience and job skills. Whatever the degree of racism, it cannot explain age differences in employment among young black males, who do not change race as they grow older.
This pattern of virtually no difference in unemployment rates between black and white teenagers when wages were determined by supply and demand in a free market, but with large and enduring racial differences in unemployment rates when minimum wage laws became effective again, also fits the economic principle that a chronic surplus of job applicants reduces the cost of discrimination to the employer.
This pattern establishes correlation between increased minimum wage rates and changing racial differences in unemployment rates, and labor force participation rates, among teenagers. If this does not conclusively prove causation, it does establish a remarkable coincidence, persisting for decades.
Alternative explanations for these changing patterns of racial differences—such as racism, poverty or inferior education among blacks—cannot establish even correlation with changing employment outcomes over the years, because all those things were worse in the first half of the twentieth century, when the unemployment rate among black teenagers in 1948 was far lower and not significantly different from the unemployment rate among white teenagers.
Severe restrictions on building homes or other structures swept through various parts of the United States during the 1970s, in the name of preserving “open space,” “saving farmland,” “protecting the environment,” “historical preservation,” and other politically attractive slogans. But, however they were characterized, what such laws and policies did in practice was forbid, or drastically reduce, the building of either housing or other structures. Coastal California, including the entire peninsula from San Francisco to San Jose, was one of the largest regions where severe building-restriction laws and policies prevailed.
The predictable effect of restricting the building of housing, as the population was growing, was a rise in housing prices, when the supply of housing was not allowed to rise as the demand rose. California home prices were very similar to those in the rest of the country before this wave of building restrictions swept across the coastal regions of the state in the 1970s. But, afterwards, home prices in coastal California rose to become some multiple of home prices in the country as a whole.40
San Francisco Bay Area home prices rose to more than three times the national average.41 In Palo Alto, adjacent to Stanford University, home prices nearly quadrupled during the 1970s, not because more expensive homes were being built—because there were no new homes built in Palo Alto during that decade. Existing homes simply skyrocketed in price.42
The racial impact of these housing restrictions was more pronounced than many racially explicit restrictions. By 2005, the black population of San Francisco was reduced to less than half of what it had been in 1970, even though the population of the city as a whole was growing.43 In an even shorter span of time, between the 1990 and 2000 censuses, three other California counties—Los Angeles County, San Mateo County, and Alameda County—had their black populations decline by more than ten thousand people each, despite increases in the general population in each of these counties.44
By contrast, Harlem was a predominantly white community as late as 1910, and there were openly proclaimed and organized efforts by white landlords and realtors to prevent blacks from moving into Harlem.45 But, like the organized white efforts to suppress black earnings in the post-bellum South, the mere presence of such organized efforts was no evidence or proof that they achieved their goal. To call such explicitly racist efforts in Harlem unsuccessful would be an understatement.
Those white landlords and realtors in Harlem who held out while others began to rent to blacks, found themselves losing white tenants who moved out of the neighborhood as blacks moved in, leaving the holdouts’ buildings with many vacancies, representing lost rent.46
No such economic consequences inhibited those residents and their elected officials in later years who restricted the building of housing in San Francisco and other coastal California communities through the political process, driving up housing prices to levels that many blacks could not afford. On the contrary, such restrictions on new building increased the market value of the existing homes of residents in those communities.
Attitudes and beliefs, however strongly held or loudly proclaimed, do not automatically translate into end results—into “what emerges”—especially when there are costs to be borne by discriminators themselves.
It may well be that the racial attitudes and beliefs held by white landlords and realtors in early twentieth-century Harlem were more hostile to blacks than the attitudes and beliefs of white residents and officials in late twentieth-century San Francisco. But, in terms of end results, the actions of the former failed to keep blacks out of Harlem, while the actions of the latter drove out of San Francisco more than half the blacks already living in that city. Costs matter.
* As a personal note, some years ago an elderly relative was crossing a busy thoroughfare in the Bronx, when she lost consciousness and fell to the ground in a high-crime neighborhood. People on the sidewalk rushed out into the street, to direct traffic around her. One of the women in the group took charge of her purse and returned it after my unconscious relative revived. Not a cent was missing from the purse.
* As a personal note, the first time I encountered a white professor at a white university with a black secretary, it was Milton Friedman at the University of Chicago in 1960—four years before the Civil Rights Act of 1964.