Printer Friendly

Racism as rent seeking.



Economists have modeled discrimination as tastes, as an optimal tariff, as a signaling device and as the result of status seeking. (1) They have not however, developed an explicitly political model of dicrimination. This omission is peculiar because race relations in the United States have been politicized since the founding of the republic, and race relations in other countries are highly politicized as well. (2)

However, since Becker's [1973] seminal work, economists have done extensive empirical work examining the impact of government intervention on ethnic minorities. (3) The existence of this body of work is testimony to the implicit recognition that the political economy of race is different from the pure economics of race. Besides intruding itself into purely economic issues, politics has often been involved in the social relations between the races as well. Of course, researchers in race relations are well aware of this fact.

The present paper makes a contribution to a political economic theory of race by modeling social rules as public goods. In contrast to most models of discrimination, the present model hypothesizes that people have preferences for other people's conformity with social norms governing race relations. The level of conformity within the general population becomes a public good. This particular public good, however, has more than the usual tendency to be underprovided, because there are several sources of inefficiency.

However, it should be noted at the outset that the discovery of this inefficiency does not lead to an argument in favor of any corrective government intervention. Rather, the demonstration of inefficiency is viewed as evidence that there are unexploited opportunities for mutually profitable exchange. Collective action may be able to generate gains that are not available through completely atomistic optimization. While collective action is notoriously costly, a Pareto-improving more is precisely one in which at least one person can be made better off without worsening anyone. Thus, an inefficient situation is one with incentives for someone to try to arrange additional transactions. Throughout the analysis, the term "efficiency" should be understood in this predictive sense, not in the usual prescriptive sense. That is, efficiency is used as a positive concept, not a normative one. The working hypothesis is that entrepreneurs organize some form of collective activity to exploit the potential gains latent in the inefficient status quo.

This discussion attempts to avoid the stark "public/private" dichotomy so frequently found in discussions of public goods. Collective action to provide social rules and public goods is a much more general phenomenon than government action to provide the same ends. And for rules governing race and ethnic relations, many forms of collective action have been used. However, the examples in this paper concern government action because government is the primary institution for organizing collective action in most modern societies. The important contrast here is between individual maximization and collective maximization, or between atomized optimization and coordinated optimization.

The model presented below contributes to understanding several unexplained puzzles. Whey has race been so consistently politicized, in spite of the widespread intuition that race-neutral law would be desirable? Why do racial differences in earnings persist, in spite of the powerful prediction that competition should erode differences between equally productive groups? Finally, why are there so many examples of government policies governing race relations that do not seem to be motivated by economic gain as usually understood? After the basic model is described, a few examples illustrate how the public goods approach can illuminate each of these issues. The next section surveys the existing approaches to discrimination and discusses their weaknesses.


Gary Becker, in his seminal work, The Economics of Discrimination, defines tastes for discrimination as a willingness on the part of one ethnic group to pay to avoid contact with members of another group. By this definition, discrimination is a costly activity. Not surprisingly, Becker's model predicts that in a competitive market, tastes for discrimination cannot support long-run wage differentials between groups that are perfect substitutes in production. (4) Employers who discriminate are at a competitive disadvantage; stated differently, the nondiscriminating employer has a comparative advantage in the ownership of capital.

Employee discrimination results in segregated workplaces, but not in wage differentials. Consumer discrimination may result in earnings differences, because the market gives people what they are willing to pay for. However, this effect is likely to be significant onl for those goods and services where the identity of the seller is known to the consumer. Moreover, the disliked ethnic group must be large relative to the number of jobs available in industries in which consumers cannot discriminate. Otherwise, occupational segregation, but not necessarily earnings differentials would exist.

One of the major objections to Becker's theory was that it failed the predictive test of positive economics, since its main conclusion of no long-run earnings differences among ethnic groups was apparently false. However, Becker's model is only a model of tastes and how these tastes operate in a purely competitive market environment. It is no exaggeration to say that the popular press and policy makers most commonly cite "tastes for discrimination" or "racism" as the primary cause of ethnic wage disparities. Becker's results is significant precisely because it shows that the effects of tastes proper are much more limited than is commonly supposed.

One of the earliest responses to Becker's work was by Anne Krueger [1963]. She argued that one could conceptualize the black and white communities as trading partners and apply optimum tariff theory from international trade. If the white community is large enough to have some market power, it can make itself better off by imposing a tariff on the hiring of black labor. Such tariffs, of course, will support long-run wage differentials.

The difficulty with Krueger's argument is that it requires some mechanism for forcing all white employers to enforce the tariff arrangement. In the absence of such an enforcement mechanism, white employers could profit by "smuggling" black labor without paying the tariff. (Imagine the difficulties of privately enforcing a tariff on foreign manufactured goods.) It is argued below that the political economy approach is useful for explaining the existence of exactly such enforcement mechanisms.

Arrow's [1973] screening theory is an alternative explanation for long-run earnings disparities. Employers use ethnicity as a proxy for productivity, and hence some groups are paid less than others. However, it is a condition of long-run screening equilibrium that prior beliefs about racial abilities must be confirmed by the evidence that employers acquire through observing the work force they hire. Thus, the screening model requires that groups differ in some relevant characteristic in the initial period. In his classic work, Arrow assumes that some groups have lower productivities. His more sophisticated analysis assumes only that one group begin the process with lower average productivity. Because of statistical discrimination of employers, these groups would tend to invest less in human capital, and hence, continue to be less productive. (5)

While it is true that groups differ in objective productivity characteristics, it is not satisfactory to assume this as an initial condition of the problem. These differences are among the phenomena to be explained. The screening model may well be an accurate description of how people behave, given differences in traits across groups and a set of stereotypes or beliefs about those traits. However, neither traits nor stereotypes are constant over time. Other factors are at work which allow stereotypes to change over time, factors which evidently outweigh the stagnation that the use of ethnicity as a screen would tend to promote. The screening theory is not compelling as a long-run explanation for group differences.

The most recent contribution to the economics of discrimination is George Akerlof's [1985] model of status-based wages. Akerlof argues that in small markets in which every participant is a potential trader, a social custom regulating hiring of various groups can be stable. A nondiscriminating entrepreneur will lose profits because the discriminating majority will boycott his business. Since markets are small by assumption, the loss of even a few potential traders will be costly in foregone profits. This line of argument provides the tariff enforcement mechanism which Krueger's argument required but did not provide.

The usefulness of this intriguing model is limited to situations which conform to its assumptions. The model is similar to Becker's consumer discrimination model, and the arguments made above about its generality apply here as well. The conditions of the model more nearly apply to rural communities in which a small group of people can monitor each other's behavior, than to an impersonal urban area.

Another important feature of Akerlof's framework is that rules governing intergroup trading must already be established, known, and accepted by nearly everyone. However, there are many instances of societies in which opinions vary as to appropriate behavior. Customs concerning race relations change over time, even in societies that once appeared to have achieved a consensus on these matters.

All of these theories seek to explain discrimination within the atomistic institutions of the market. None of them attempts to explain discrimination as the orchestrated result of collective action. The present work attempts to do exactly that. Note at the outset that the processes described here are equally applicable to preferences for integration as to preferences for segregation.


The Structure of Preferences

Consider some social norm or custom. Individuals in society gain utility from conforming to the norm themselves, and from tne general level of conformity in the population. Individual i's utility is given by

[U.sub.i] = [U.sub.i]([x.sub.i], [c.sub.i]; [alpha])

The composite commodity consumed by individual i is [x.sub.i], Let [c.sub.i] be a dichotomous variable indicating individual i's conformity to the norm; i.e.,

[c.sub.] = 1, if individual i conforms;

0, otherwise. (6)

The variable [alpha] represents the fraction of the population that accepts the norm or conforms to it. Individuals gain utility from knowing that other people are conforming to the norm. (7) The utility gain may be a gain in pure psychic income, if, for example, employers gain satisfaction from knowing that blacks are being kept properly "in their place." Or the utility increase may arise from an increase in money income if universal observance of the rule increases profits. Rules governing race relations can be thought of as public goods. Everyone "consumes" the same set of social norms. The extent to which these norms are observed in a collectively consumed, nonexcludable good.

The acceptance rate, [alpha], is determined in the following way. Let [p.sub.c] be the market price of conforming to the norm. A person accepts the norm if the price of conforming is less than the marginal rate of substitution between market goods and conformity, or if

[p.sub.c] [is less than] [U.sub.c]/[U.sub.x] * [delta].

Note that Becker's concept of "tastes for discrimination" is a special case of [delta]. Whether a person conforms or not depends only on the price of conforming. Thus,

[Mathematical Expression Omitted]

where (N+1) is the total number of individuals in society.

In the analysis that follows, utility is assumed to be increasing in all its arguments. The society universally agrees that the norm is good, in that everyone gains utility both from conforming and from seeing other people conform. The only area of disagreement is the extent to which people value the norm. (8) The model could be expanded to consider the possibility that some people actively resist the norm. For them, the norm is a bad, and they would not conform at any positive price. They will only conform if they are compensated. It is doubtful that this refinement would substantially change the basic results of the model.

This formulation of rules or norms is quite general. The norm could be to hire only white workers. The price of conforming to this norm would be the wage differential between the races. So this formulation of norms includes the standard economic questions considered by Becker, such as whom to hire, work with and purchase from. The norm could also be obeying cartel rules, which many members have an incentive to violate. Or the norm could govern behavior that is not usually considered economic, such as social relations between races and ethnic groups. The norm could be not sitting next to a black person on a streetcar or bus. The cost of conforming would be low if the car were nearly empty, but higher if the alternative were to stand in the aisle the entire trip.

Note also that the traditional economic approach considers only the individual's preferences for his own behavior. The model here specifically considers the individual's preferences for other people's behavior. People gain utility from the widespread acceptance of the norm. But this utility is different from the utility they gain from conforming themselves. That is, [U.sub.[alpha]]/=U.sub.x] * [alpha] is in general diferent from [delta]. If [alpha] [is not greater than] [delta], people are willing to pay something to see other people obey a rule that they would prefer to violate themselves. This is the situation of the cartel "cheat." On the other hand, if [alpha] [is less than] [delta], people are willing to pay more to conform to a rule than to pay to see someone else obey the rule. An individual may have intense preferences for marrying someone of his own race, but be relatively indifferent about other people's marriage choices.

The Production of Conformity

It is obvious from the formulation of the utility function that individuals have preferences for the behavior of other people. What methods are available for changing other people's behavior? It is assumed that individuals can expend resources, r, to lower the price of conforming. These resources can be conceptualized as voluntary contributions to public enforcement activities, or as the opportunity cost of personal enforcement efforts. Individuals may vary in their ability to lower [p.sub.c]; [dp.sub.c]/[dr.sub.i] = [[Gamma].sub.i] is the change in [p.sub.c] generated by expenditures of the ith individual. It is assumed that [Gamma] is negative for all i; efficient or low-cost producers of conformity are those individuals with relatively large absolute values of [[Gamma.sub.i].

Thus, the price of conforming depends on the contributions made by all the individuals in society. That is,

[P.sub.c] = [P.sub.c]([r.sub.0], ..., [r.sub.N]).

[alpha] in turn depends on [p.sub.c], so

[alpha] = [alpha][[p.sub.c]([r.sub.0], ..., [r.sub.N])].

For simplicity, it is assumed that each [[Gamma].sub.i] is independent of each [[Gamma].sub.i] for i [is not equal to] j.

This framework deals with the enforcement of the norm in a particularly simple way. Neither public nor private parties enforce the norm in the usual sense. Instead, resources are expended to lower the price of conforming. Once the price is set, each individual makes his personal calculation as to whether to conform or not. Thus, the problems of detecting and sanctioning violations are incorporated into the more general parameter of the price of conforming.

The analysis that follows focuses on the comparison between atomistic and coordinated provision of conformity. No assumption is made about the relative efficiencies of private and public provision of the public good, in order to focus more directly on the coordination problem of providing this particular public good. Many interesting and plausible assumptions about the relative efficiencies of the public and private sectors could be made. For instance, people may be more inclined to obey a law than a similar social custom. On the other hand, some kinds of behavior may be more easily monitored by private parties than by officials of the government. Public provision would be technologically inefficient, but there might still exist gains from some kind of coordinated action. The following sections demonstrate the existence of gains from coordinated action.

Atomistic Provision of Conformity

The problem for the representative individual is to maximize utility subject to the budget constraint.

maxL = [U.sub.0]([x.sub.0], [c.sub.0]; [alpha]([r.sub.0], ...[r.sub.N])) +

[mu] [I-[x.sub.0] - [c.sub.0][P.sub.c]-[r.sub.0]].

The first-order conditions on [x.sub.0] and [c.sub.0] imply that [mu] =[U.sub.x] = [U.sub.c]/[P.sub.c]. This implies the conformity condition alluded to above, namely the individual conforms (i.e., c = 1) if [p.sub.c] [is less than] [U.sub.c]/[U.sub.x] = [delta], and does not conform otherwise.

These conditions, combined with the first order condition on [r.sub.0], yield the following:

[U.sup.0.sub.[alpha] / U.sup.0.sub.x = [1 + [Gamma].sub.0.c.sub.0 + p.sub.c.(dc.sub.0./dp.sub.c.)[Gamma].sub.0.] / [Gamma].sub.0.(d[alpha]/dp.sub.c.)].

The right side of (1) tells the market trade-off between x and [alpha], while the left side tells the internal trade-off for individual 0. The denominator of the right side tells the [alpha] foregone if consumer 0 spends one dollar on market goods. This can be termed the price of x in terms of [alpha]. The numerator tells the x foregone if one dollar is spent on [alpha], or the price of [alpha] in terms of x. To see this, note that an expenditure on [alpha] has several effects. First, the consumer has one less dollar to spend on x; this is captured by the first term of unity. Second, the expenditure on [alpha] produces a marginally lower price of conforming. If the individual is already conforming (i.e., if c=1), the second term in the numerator is negative, and the net cost of increasing acceptance is less than it otherwise would be. In effect, lowering [p.sub.c] creates a rent for those who conform to the norm. For those who do not conform, of course, c=0, there are no rents, and the second term vanishes. Finally, lower [p.sub.c] will induce conformity on the part of those on the margin. The last term captures this effect.

Coordinated Provision of Conformity

This section shows that the atomistic solution does not satisfy the Pareto efficiency criterion. The utility of the representative individual is maximized subject to the resource constraint and the utility levels of all the other individuals.

maxL = [U.sub.0]{[x.sub.0], [c.sub.0]; [alpha][[p.sub.c]([r.sub.0], ...[r.sub.N])]}

[Mathematical Expression Omitted]

[Mathematical Expression Omitted]

where Y is national income.

Not surprisingly, the first-order condition for the public good is quite different from (1).

[Mathematical Expression Omitted]

The standard public goods result requires that the marginal rate of transformation in production be equated to the sum of the marginal rates of substitution, rather than to the individual's own marginal rate of substitution. The standard result is illustrated here by the fact that the left side of (3) is the sum of the marginal rates of substitution.

However, the right side of (3) differs from the right side of (1), because the social marginal rate of transformation (equation (3)) differs from the marginal rate of transformation as perceived by the individual (equation (1)). Lowering [p.sub.c] generates rents to everyone who conforms, not just to the individual himself. A comparison of the second term in equations (1) and (3) illustrates this point: in (3), the term sums the rents over all the individuals who conform. In addition, the third term of (3) takes account of the increase in the number of individuals who conform. Whereas the atomized solution only considered the effect of shifting the individual agent off the margin of acceptance, the coordinated solution incorporates the fact that an individual's expenditure to lower the price may induce many people to conform who otherwise would not have.

Once rules are modeled as public goods, it is not surprising to find that rules are underprovided. It is somewhat more surprising to find the additional inefficiencies represented by the second and third terms of equation (3). The importance of all these inefficiencies is that they represent potential gains from coordinating the provision of conformity. If one individual increased r, the price of conforming would fall enough to benefit the inframarginal conformers, even if they agreed to pay r to the individual whose expenditure increased. Even remaining completely agnostic on the question of whether government or private provision is more technologically efficient, the inefficiencies illustrated in (3) represent benefits that could be obtained from some kind of collective action, be it public or private.

It is worthwhile at this point to compare the present model of conforming to norms or custom with Akerlof's model. In Akerlof's model, the cost of conforming to the norm is the loss in productive efficiency. The presumption is that initially everyone accepts the norm, or that [delta] [is greater than] [p.sub.c]. If a new person enters the market and attempts to violate the norm, the existing members impose sanctions. But the sanctions are costless to the individual imposing them. Individuals trade identical commodities in a thin market. Because the market is thin, the loss of even a few trading partners imposes costs on the nonconformists. However, the individual who refuses to trade loses nothing. By assumption, he can trade with any of the numerous conformers. But more to the point, the individuals are trading identical commodities, so there are no gains from trade to forego. Hence, by refusing to trade, people are lowering the cost of conforming. But they can lower [p.sub.c] at zero cost to themselves; i.e., r=0 and [Gamma] is infinite. It is not surprising that the norm is self-enforcing under these circumstances.

In addition to the inefficiencies on the demand side shown in equation (3), inefficiencies arise on the production side as well. These arise because people differ in their ability to reduce the cost of conforming; i.e., the [[Gamma].sub.i]'s differ. A "high-cost" producer of conformity (i.e., one with a low [Gamma]) could give a dollar to a "low-cost" producer and lower [p.sub.c] by more than if he had spent the dollar himself. These trades, of course, are not accounted for in the atomized maximization process.

To see this, recall that (3) was derived from the first-order condition on [r.sub.0]. There is a similar condition for every [r.sub.i]. That is,

[Mathematical Expression Omitted]

Therefore, the right side of (3) must be equated to the right side of (4) to achieve production efficiency. Otherwise, there remain unexploited gains from trade between low and high [Gamma] individuals. This condition does not usually arise in public goods problems because the contributions of individuals are assumed to be interchangeable. But the present analysis allows the possibility that some people are more efficient at lowering the cost of conformity than others. The low-cost producer may be the county sheriff, the tribal elder, the shop foreman, the town gossip, or the local clergyman. Anyone who can influence people's behavior at relatively low cost to him or herself could be considered a low cost producer.

Thus, there are several sources of inefficiency in the atomistic provision of social norms. Economists typically argue that such inefficiencies provide a rationale for government provision of the public good, or at least, government intervention to correct the distorted price signals that generate the externalities. This paper should not be interpreted as advocating government provision of social norms; quite the contrary. The point is not normative, but positive. The model is not intended to prescribe, as much as to predict.

The inefficiencies arising from atomistic provision of social norms create an incentive for coordinated action. In societies where the government is the major institution available for providing collective action and correcting externalities, the model predicts that government will be drawn into the business of enforcing social norms. In societies with other means of coordinating the activities of individuals, social norms will be enforced by these institutions.


Rent seeking can occur any time a change in the rule structure is anticipated. The change in the rules may either enhance or inhibit efficiency. One typically thinks of rent seeking as expending resources to obtain the right to engage in a producing activity that is inefficient for society as a whole; import quotas are the paradigmatic example. Inefficiencies are created both by the import quota and the efforts to obtain the quota. But people can seek rents, even when the rents arise from an activity that ordinarily is efficiency improving. For example, moving from common property to private property need not increase overall efficiency because competition to obtain the private right can dissipate the usual gains (see Anderson and Hill [1983]).

With this in mind, return to the problem of enforcing racial norms. It has been shown above that atomistic provision of norms governing race relations will not be efficient in the usual sense. This means that there are unexploited opportunities for trade among individuals. For example, two individuals with different values of [delta] can be made better off if the high [delta] individual pays the low [delta] individual to refrain from social relations with people of other races. Similarly, two people with different values of [Gamma] can be made better off, regardless of their preferences, if the low [Gamma] individual gives a dollar to the high [Gamma] individual to spend on lowering the price of conformity.

There are potential gains (or rents) to be obtained from moving from an atomistic regime to a regime that has at least as much coordination as involved in side payments. But shifting regimes is generally not costless. Resources have to be expended to organize a coordinated solution, whether the coordination involves creating an enforcement bureaucracy, or establishing a police state, or nothing more elaborate than making side payments. These resources are the rent-seeking costs.

Embedded within the inefficiencies are potential rewards to political and social entrepreneurs who come up with some collective mechanism for enforcing social norms. Inefficiencies constitute foregone gains from trade that perform a function similar to foregone profit opportunities in price theory; they lure entrepreneurs into a profitable activity. It is possible that the gains from coordinated action are sufficient to pay off the losers, provide a reward to the entrepreneur, and still leave the rest of society just as well off. And if part of a political program involves disenfranchising or otherwise disabling the losers, it may not even be necessary to pay them off. In this case, of course, the potential gains to the entrepreneur are even greater.

What does all of this have to do with racism as ordinarily understood? Racism is often thought of as simply preferences for members of one's own group. In an atomistic world, preferences for general conformity to one's views of correct behavior may not be expressed. These preferences may be latent because the collective action problem is too severe. There are rents to be made in trying to satisfy these preferences. But earning these rents allows the latent preferences to surface. Preferences can matter, but not in the strictly economic sphere analyzed by Becker [1973]. Preferences may matter in the collective in a way that they do not matter individually. Thus, at the very least, racism can emerge from the rent-seeking process.

At its worst, however, the rent-seeking process can actually intensify the already existing preferences. In effect, people can make themselves better off by drawing rings around themselves, by defining themselves and their group more sharply. These gains can arise strictly from preferences, and not necessarily from the usual sorts of economic gains. However, it is certainly possible to increase the economic well-being of one group by effectively enforcing an economic cartel arrangement among its members. There are numerous examples of both phenomena, rules that enforce preferences and rules that enforce cartels. A few of these examples are cited below.

Psychic Rents and The Sen Paradox

Before turning to specific examples, note the relation between the present analysis and the Sen paradox. The present formulation of preferences for race relations illustrates the Sen [1970] paradox: the Pareto optimal outcome is not liberal in the usual meaning of the term. The illiberal outcome of segregation can occur because of the twin properties of the publicness characteristic of the good on the one hand and the availability of political power on the other hand.

And, similar to Sen's analysis, the analysis here suggests that the Pareto criterion is not adequate for the formation of public policy. That is, the Pareto criterion cannot discriminate between fully private preferences and meddlesome preferences. Thus, the Pareto apparatus does not have its usual appealing normative properties when applied to race relations. We must look elsewhere for guidance in policy making.

The basic difficulty stems from the peculiar elements of the utility function in both the present analysis and in Sen's. The economist's fondness for the Pareto criterion is based on an implicit value on giving people what they want. In many cases, the efficiency approach can be used without requiring the economist or anyone else to pass judgment upon the substance of people's preferences. For this reason, the efficiency standard leads to a "live and let live," liberal approach to preferences, with little obvious conflict among individuals.

But if people want to place constraints on the consumption patterns or social activities of their neighbors, a blind application of the Pareto apparatus will lead to paradoxical results. An efficient outcome is no longer the liberal outcome that allows people to do what they want. Neither the economist nor the public can remain agnostic about the content of people's preferences when those preferences include a desire to deny options to other people. The efficiency criterion must be supplemented by other normative criteria. Instead of using efficiency because it leads to liberal individualism, perhaps liberal individualism itself should be the criterion.

However, the use of the familiar efficiency apparatus does contain a useful positive insight into race relations, despite these normative deficiencies. The Pareto apparatus explains why race is likely to be politicized. In most modern societies, the government is the institution most likely to be used to provide the coordination necessary to overcome inefficiencies from atomistic action. This accounts for the persistent politicization of race, despite widespread recognition of the social problems it causes. To the extent potential gains exist, there will be rents to be earned by becoming a political entrepreneur who politicizes race.


This section presents examples of government actions that are difficult to categorize as the result of pure economic rent seeking. At the same time, these examples illustrate the possibility of using the government to provide enforcement for a social norm that had not been universally observed.

Streetcar Segregation

Race relations in the U.S. South were in a state of flux in the period immediately after the Civil War and remained so until the turn of the century, when the "Jim Crow" system of rigid segregation was put into place. Segregation was supported by both law and custom by the time it was dismantled during the Civil Rights movement of the 1960s.

Municipal streetcars were segregated by law in many Southern cities. Roback's [1986] investigation of the origins of streetcar segregation shows that legislation requiring segregation did not simply codify existing practices. Prior to legislation, streetcar passengers in many cities sat wherever they wished, and next to anyone they wished. There is little to indicate that segregation was introduced in response to the demands of passengers, and in some cases, passengers of both races were dissatisfied with the new rule. Moreover, some of the streetcar companies themselves actively resisted segregation, on the grounds that segregation would be too expensive.

White passengers seemed to be indifferent about segregation; streetcar companies resisted segregation; certainly black passengers resisted segregation. Who then wanted it badly enough to work for its introduction? The most likely candidates are politicians who believed that there existed latent sentiment in favor of segregation among whites. Political entrepreneurs could offer white voters something they valued enough to vote for, but not enough to bear the costs privately. Through collective action, the costs of segregation could be imposed on the (disenfranchised) black passengers and the (regulated) streetcar companies.

Berea College

Berea College is a private religious school in Kentucky that was integrated at its founding prior to the Civil War. In 1904, the state legislature outlawed interracial education, even in private schools. (Kentucky public schools were already segregated.) This law was widely known to be directed exclusively at Berea College. The school had operated successfully as an integrated institution for forty years and the wrath of the community had been unable to stop it.

The economic motives for forcing Berea to segregate appear to have been minimal. That is, there were no obvious resources transferred from black to white hands by this legislation. This is in direct contrast with the school segregation legislation discussed below, which was clearly aimed at resource transfer. Evidently, the major motivation in the Berea case was the desire of lawmakers to associate themselves with the widespread demand for segregation in a manner that was relatively inexpensive for themselves and their constituents. (9)

Interracial Marriage

The frontier societies of European countries in the New World and in Africa did not initially have taboos against interracial marriage. Marriage or concubinage with natives or with nonwhite slave women was common in South Africa, and not unheard of in colonial Virginia. As the taboos took shape in Virginia, the strongest sanctions were imposed on unions between white women and non-Europeans. This policy was clearly an attempt to allocate the scarce resource of European women toward white men and away from black and Native American men. Evidently, racism alone was not enough to prevent all women from choosing nonwhite partners. In South Africa, interracial marriage was common into the late nineteenth century. Both societies eventually outlawed miscegenation (see Fredrickson [1982]).


The example of interracial marriage links psychic rent seeking with pure economic rent seeking. The ban on interracial marriage both enforced preferences and allocated scarce resources, namely European women, to the politically dominant group. This section discusses economic rent seeking more generally.

In the previous discussion, it was assumed that people had preferences for discrimination or for segregation. Rents arose in that context when people were able to satisfy their subjective preferences for less than their willingness to pay. Different assumptions are made in the following discussion of economic rent seeking. In particular, it is assumed that people have no particular preferences for discrimination or segregation. This assumption allows isolation of the pure effects of racism from the exclusively monetary benefits of familiar economic rent seeking, realizing that most historical examples combine elements of both discriminatory preferences and economic gain.

Ethnic groups can be powerful rent-seeking bodies because the identities of the winners and losers are clear from the outset. People cannot readily change their ethnic identity, making many race-specific transfers difficult to evade. That is, a black person cannot evade a restriction on blacks owning land by becoming not black, although a potential railroad competitor could evade an entry restriction by providing an alternative means of transportation. More importantly, those who initiate race-specific legislation know that they will never be members of the dispossessed group, and hence they also know that they will never be one of the direct victims of the legislation. Few other rent-seeking activities can offer such a guarantee. For example, people who vote for income redistribution are at least partially restrained by the knowledge that their own income may change so that they may become net losers in the redistribution game. (10)

Ethnic groups are in many ways paradigmatic Olson [1982; 1971] interest groups. Their members are easily identified. Ethnic groups are typically organized for social, religious and cultural purposes. These organizational structures can be turned toward political objectives as well. All of these factors operate toward making it possible for the leaders of ethnic groups to make ethnicity a politically "salient" issue, to borrow the terminology of Rabushka and Shepsle [1972, 55].

Thus, even if groups begin without any particular race consciousness or ethnic animosity, it may be worthwhile to create such consciousness to enhance the economic gains that can be won for the members of the group. Professor W.H. Hutt has long made this argument for the case of South Africa. Hutt wrote in his 1964 classic work, The Economics of the Colour Bar,

"We do not, however, find in colour prejudice as such the main origin--nor, perhaps, even the most important cause--of most economic colour bars. The chief source of colour discrimination is, I suggest, to be found in the natural determination to defend economic privilege,... non-whites simply happening to be the essentially underprivileged groups in South Africa." (11)

As in the case of social rules, the rents arise because of the publicness of the good and the availability of political power. Cartel gains are a public good within any cartel, and thus, the rents accruing to an ethnic cartel are public goods. Political power is often useful both in obtaining economic rents and in preventing encroachment on those rents by outsiders. Rules need to be enforced, whether the rules are social norms or cartel agreements. Organizing ethnic groups into economic cartels is a promising field for political entrepreneurs.

Exclusion of ethnic groups from economic activities such as occupations, education or land ownership is a typical device for generating rents for the dominant ethnic groups. This is particularly true in countries which have constitutional prohibitions against "class legislation" but which do allow for a variety of economic regulations that are ostensibly race-neutral.

Exclusion From Occupations: The South African Colour Bar

The standard price theoretic framework can be employed to analyze these exclusions. Excluding an ethnic group from some occupation reduces the supply of labor to that occupation, thus bidding up its wage. At the same time, the members of the excluded ethnic group will work in their next best alternative employment. This increases the supply of labor to that occupation, depressing its wage rate.

Exclusion can be used by the majority group labor force as a means of reserving certain occupations for themselves and keeping the wages high. Alternatively, majority capital owners or employers might seek means of keeping the wages of their minority labor force low. Limiting the alternative opportunities of minority labor will have that effect.

Exclusion restrictions offer the cartel enforcement mechanisms required by Krueger's [1963] tariff model. The politically dominant ethnic group behaves as a labor market cartel, using the government as an enforcer. Employer cartels designed to keep minority wages low can be broken by employers willing to pay higher wages. Limiting the alternative employment opportunities for minorities reduces the equilibrium wage, meeting the objective of the cartel without the associated enforcement costs.

One of the cornerstones of South African apartheid is the Mine and Works Act of 1911, also called the Colour Bar Act. This legislation reserved the skilled mining jobs for Europeans. In addition, the "civilised labour policy" reserved jobs for whites by requiring a minimum wage. The wage was set high enough to eliminate the competitive advantage that Africans had by working at wages lower than Europeans would accept. This policy was initiated in 1926, after the election of the "Pact Government" in 1924, a coalition of English-speaking trade unionists and rural Afrikaaner nationalists. (12)

These laws are straightforward examples of economic rent seeking. White employers had been willing to hire African workers for a wide variety of mining and industrial jobs, including skilled jobs. Africans were competing successfully with European workers. The white trade unions obtained legislation to limit the competition they faced from native workers.

Human Capital Restrictions: Education in the U.S. South

Besides occupational exclusion, the majority may exclude the minority from the acquisition of capital or other factors of production. Educational exclusion, for example, may be a method of limiting entry to certain occupations. The same groups discussed above as likely to favor occupational exclusion may favor educational restrictions as well.

Economists have long been aware of the important consequences of governmental discrimination in the provision of education. Differential acquisition of human capital can have long-run implications for differences in earnings across races. Several economists have documented the link between income differentials and human capital differentials across races. These economists have pointed out the less competitive nature of the public school market, in contrast to private markets. Thus, tastes for discrimination by school officials can persist longer than might be expected in a competitive market. (13)

Moreover, the public choice aspect of this problem was noted in the thirties by the noted black historian, Horace Mann Bond [1934]. Bond argues that school funds in Southern states were allocated to the counties on the basis of total school-aged children in the county. The counties themselves then allocated the state funds according to their own criteria. Bond argues that the black belt counties, i.e., counties with relatively large black populations, distributed the money allocated for black children to the white schools. School boards were controlled by whites and black voters were disenfranchised. Thus, white schools in black belt counties were the best funded schools in the South. (14)

Bond's argument has recently been studied in some detail by economists. Smith [1984] alludes to it and Margo [1985] explicitly tests it. Margo finds that higher proportion of black students is indeed correlated with higher expenditures on white education, and lower expenditures on black education. He also finds that these effects were decidedly stronger after black voters were disenfranchised.

Land Restrictions: The South African Land Acts

Restrictions on access to land is another means of extracting rents from the minority for the majority. Such restrictions allow the majority access to land at a lower price than would otherwise be available. In addition, these restrictions limit the agricultural opportunities of the minority groups, thus making them available to work in industrial employment at reduced wages.

South African apartheid includes restrictions on land ownership by blacks, in addition to the occupational restrictions generated by the Colour Bar. The Land Acts of 1913 allotted about 13 percent of the country as "reserves" or "homelands" for the African majority (see Fredrickson [1982, 244]). This restriction provided two different types of rents to the white community: more land was available to white farmers at lower price than would otherwise be the case, and the land allotted to blacks was insufficient to support them as independent farmers or hunters. By limiting agricultural opportunities for the native workers, the Land Acts increased the supply of African labor to white employers, both agricultural and industrial employers. (15) Indeed the advocates of the Land Acts were quite explicit in their intentions. According to one prominent South African politician of the 1930s and 1940s, "it is inconceivable that the white man should be able completely to dispense with the black man's labor on his farms, in his mines, in his factories; it is just as inconceivable that there should be set aside for the black man's occupation land sufficient to provide for all his needs independent of the white man's wages." (16)


As mentioned at the outset, much of the analysis here applies equally to tastes for integration and equality as to tastes for segregation and oppression. State power can be used to transfer income from the majority to the minority or vice versa. Government enforcement can create a social norm of indifferent mingling as well as a norm of strict separation.

It should be said at this point that this analysis differs sharply from the position that has come to be associated with many conservatives. Some conservatives argue with current civil rights policy on the grounds that the policies do not in fact help minorities. But surely one could devise some policy that would actually be successful in transferring income to minorities. That is why this line of argument is infuriating to those who do not share the basic world view of the conservatives, and ultimately unsatisfying even to those who do. The law is not ineffectual in changing interracial behavior patterns or even interracial attitudes. On the contrary, the law, while certainly not omnipotent, can be a powerful force in shaping race relations, so powerful in fact, that the legal system can do what cannot be done privately.

The power of the state to create customs is a two-edged sword. It can and will be used against the currently favored group, if the political power coalitions shift. An argument in favor of expanding state power to intervene on behalf of particular ethnic groups carries with it the risk that the enlarged powers will one day be turned against those very ethnic groups. (17)

How then should the role of the state in race relations be evaluated? Social norms are public goods. But this paper does not argue in favor of state determination and enforcement of such norms: quite the contrary. The analysis here is intended as a positive account for why race relations are so consistently the subject of political wrangling. The publicness of social norms means that opportunities exist for political entrepreneurship. In the absence of some explicit constitutional prohibitions or sanctions, race relations are bound to be politicized, in one direction or the other.

The Founding Fathers had all too much experience with religious warfare. Because of that experience, they created specific safeguards for religious freedom and autonomy. In short, they placed constitutional limitations on the politicization of religion. If they had not, the examples of this paper might well have been about Catholics and Protestants, rather than about blacks and whites.

Race-neutral law would be far more desirable than a system that can be bent to favor some groups at the expense of others. It is an open question whether it is actually possible to devise a set of constitutional restraints that would provide an enduring system of color-blind law. But it is the correct question to ask.

(1) As in Becker [1973], Krueger [1963], Arrow [1973], and Akerlof [1985].

(2) It should be noted that political scientists have developed models of the politicization of race. Two notable examples include Rabushka and Shepsle [1972] and Rabushka [1974].

(3) For evidence from the nineteenth century, see Higgs [1980] and DeCanio [1974]. For early twentieth-century agricultural policies, see Wright [1986]. See also Alston's excellent review essay of Wright's book [1987]. For modern treatments of educational policy, see Borcherding [1977] as well as the references in note 13 below. In addition to agricultural and educational policies, economists have studied the effects of numerous other government policies on ethnic minorities. For example, Landes [1980] argues that immigrant women were disproportionately affected by maximum hours legislation.

(4) The requirements for this generalization are that there must be a sufficient number of nondiscriminating employers so that the labor market remains competitive, the marginal return to the entrepreneurial factor must not decline too rapidly, and the costs of adjustment must be relatively small.

(5) See Arrow [1973, 26-31]. Also see Lundberg and Startz [1983] for a more recent variation of the screening theory.

(6) Conformity could be defined as a continuous variable, describing the number of times the person conforms as a fraction of the total possible opportunities for conforming. The present dichotomous definition is sufficiently general to illustrate the major points of the argument.

(7) Note the relationship between this formulation and Sen's [1970] concept of "meddlesome preferences."

(8) The differences among people are characterized as differences in their preferences. It is assumed that everyone faces the same market cost of the behavior.

(9) For more details, see Roback [1990]. Interestingly, the college used private property and voluntary association as their key arguments in the court challenges.

(10) Higgs has made some of these points see [1980] especially pages 10-11 and 123-34. See also Sowell [1983, 16-7].

(11) See Hutt [1964]. For a discussion of Hutt's approach to segregation, see Roback, [1988].

(12) See Fredrickson [1982, 233]. See also Hutt [1964, 62-80]. For an econometric treatment of some of these issues, see Anton D. Lowenberg [1989].

(13) On the link between incomes and schooling, see Welch [1974], Smith and Welch [1977], Smith [1984], and Margo [1986]. Margo [1985] stresses that school board officials, while less constrained than private entrepreneurs, are partially restrained by the mobility of black families.

(14) Bond's argument is echoed by other historians. See, for example, Harlan [1958] and Bullock [1967].

(15) This is a large part of "The Rationale of Apartheid" as ably demonstrated by Mats Kundahl in his 1982 article of that name. For similar arguments for the West Indian case, see Eric Foner [1983].

(16) Jan H. Hofmeyr quoted in Fredrickson [1982, 244]. Fredrickson describes Hofmeyr as a "relatively liberal politician who was to become deputy prime minister in the 1940s."

(17) The argument is similar to that made by James Buchanan [1986].


Akerlof, George A. "Discriminatory, Status-Based Wages among Tradition-Oriented, Stochastically Trading Coconut Producers." Journal of Political Economy, April 1985, 265-76.

Alston, Lee. "The Wright Interpretation of Southern U.S. Economic Development: A Review Essay of Old South, New South, by Gavin Wright." Agricultural History, Fall 1987, 52-67.

Anderson, Terry and Peter J. Hill. "Privatizing the Commons: An Improvement?" Southern Economic Journal, October 1983, 438-50.

Arrow, Kenneth J. "The Theory of Discrimination," in Discrimination in Labor Markets, edited by Orley Ashenfelter and Albert Rees. Princeton: Princeton University Press, 1973.

Becker, Gary S. The Economics of Discrimination, 2nd Edition. Chicago: University of Chicago Press, 1973. (First edition published 1957.)

Bond, Horace Mann. The Education of the Negro in the American Social Order. New York: Prentice Hall, 1934.

Borcherding, Thomas. "An Economic Approach to School Integration: Public Choice with Tie-ins." Public Choice, Fall 1977, 53-77.

Buchanan, James. "Politics and Meddlesome Preferences," in Smoking and Society, edited by Robert D. Tollison. Lexington: Lexington Books, 1986.

Bullock, Henry Allen. A History of Negro Education in the South from 1619 to the Present. Cambridge: Harvard University Press, 1967.

DeCanio, Stephen. Agriculture in the Postbellum South: The Economics of Production and Supply. Cambridge: MIT Press, 1974.

Foner, Eric. Nothing but Freedom: Emancipation and its Legacy. Baton Rouge: Louisiana State University Press, 1983.

Fredrickson, George M. White Supremacy: A Comparative Study in American and South African History. New York: Oxford University Press, 1982.

Harlan, Louis R. Separate and Unequal: Public School Campaigns and Racism in the Southern Seaboard States, 1901-1915. Chapel Hill: University of North Carolina Press, 1958.

Higgs, Robert. Competition and Coercion: Blacks in the American Economy, 1865-1914. Chicago: University of Chicago Press, 1980.

Hutt, W. H. The Economics of the Colour Bar: A Study of the Economic Origins and Consequences of Racial Segregation. London: Andre Deutsch, Ltd., 1964.

Krueger, Anne. "Economics of Discrimination." Journal of Political Economy, October 1963, 481-86.

Kundahl, Mats. "The Rationale of Apartheid." American Economic Review, December 1982, 1169-79.

Landes, Elisabeth. "The Effect of State Maximum-Hours Laws on the Employment of Women in 1920." Journal of Political Economy, June 1980, 476-94.

Lowenberg, Anton D. "An Economic Theory of Apartheid." Economic Inquiry, January 1989, 57-74.

Lundberg, Shelly J. and Richard Startz. "Private Discrimination and Social Intervention in Competitive Labor Markets." American Economic Review, June 1983, 340-47.

Margo, Robert A. "Race, Educational Attainment and the 1940 Census." Journal of Economic History, March 1986, 189-98.

Margo, Robert A. Disfranchisement, School Finance, and the Economics of Segregated Schools in the United States South, 1890-1910. New York: Garland Publishing, Inc. 1985.

Olson, Mancur. The Rise and Decline of Nations: Economic Growth, Stagflation and Social Rigidities. New Haven: Yale University Press, 1982.

Olson, Mancur. The Logic of Collective Action. Cambridge: Harvard University Press, 1971.

Rabushka, Alvin. A Theory of Racial Harmony. Columbia: University of South Carolina Press, 1974.

Rabushka, Alvin and Kenneth A. Shepsle. Politics in Plural Societies: A Theory of Democratic Instability. Columbus: Merrill, 1972.

Roback, Jennifer. "The Political Economy of Segregation: The Case of Segregated Streetcars." Journal of Economic History, December 1986, 893-918.

Roback, Jennifer. "W. H. Hutt's The Economics of the Colour Bar." Managerial and Decision Economics, Winter 1988, 65-70.

Roback, Jennifer. "Rules vs. Discretion: Berea College v. Kentucky," International Journal of Group Tensions, forthcoming 1990.

Sen, Amartya. "The Impossibility of a Paretian Liberal." Journal of Political Economy, January 1970,.

Smith, James P. "Race and Human Capital." American Economic Review, September 1984, 685-98.

Smith, James P. and Finis Welch. "Black-White Male Wage Ratios: 1960-1970." American Economic Review, June 1977, 323-38.

Sowell, Thomas. The Economics and Politics of Race: An International Perspective. New York: Morrow and Company, 1983.

Welch, Finis. "Education and Racial Discrimination," in Discrimination in Labor Markets, edited by Orley Ashenfelter and Albert Rees. Princeton: Princeton University Press, 1974.

Wright, Gavin. Old South, New South: Revolutions in the Southern Economy Since the Civil War. New York: Basic Books, 1986.
COPYRIGHT 1989 Western Economic Association International
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Roback, Jennifer
Publication:Economic Inquiry
Date:Oct 1, 1989
Previous Article:Parental altruism and self-interest: child labor among late nineteenth-century American families.
Next Article:Industrial structure and monopoly power in the federal bureaucracy: an empirical analysis.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters