Printer Friendly

An economic theory of apartheid.


Apartheid is a regulatory system designed to effect redistributions in favor of white workers and farmers at the expense of black workers and white capitalists. This paper uses a competitive interest group theory of the apartheid state to formalize a collective choice analysis of apartheid as endogenous policy. The "level" of apartheid is conceived as a continuous variable that is determined by the relative influence of competing interest groups within the white polity and by the costs of maintaining and defending apartheid institutions. Some empirical implications of this approach are explored.


Since South African apartheid embodies essentially a complex web of politically determined restrictions affecting the level of employment and mobility of the black labor force, it is not surprising that most economic analyses of apartheid institutions have focused on their evident allocative inefficiency and distributional inequity (see Nattrass [1981, 31-32]). Some economists have hinted at the possibility of formally modeling the economic, social and political institutions of apartheid as endogenous products of a process of collective choice. In other words, the question has been raised of the existence of an economic "rationale" for apartheid. Thus, Richard Porter [1978] and subsequently Mats Lundahl [1982] have initiated rigorous analysis of what Porter terms a "South African-type" economy--a competitive market system, the operation of which is constrained by state-imposed apartheid regulation.

While the Porter-Lundahl characterization of the South African-type economy is an invaluable starting point for an economic analysis of apartheid, it falls short of providing a satisfactory explanation for the existence and extent of South African discriminatory policies. The inevitable contradiction between economic efficiency and the political requirements of apartheid leads both Porter and Lundahl to treat the goals of apartheid as exogenously determined by political considerations.(1) The purpose of this paper is to show how it is possible to provide an explicitly economic explanation of apartheid institutions. Apartheid is viewed here as a vector of policies which can be varied in intensity along a continuum. Thus the level of apartheid regulation is treated as a continuous endogenous variable, responsive to the relative influences of interest groups which receive benefits or incur costs associated with apartheid.

The intention is not to claim originality in recognizing the economic motives for apartheid policy (which have been well documented in the literature, as indicated in section II), but rather to formalize this approach (in section III) and to suggest some methods of testing its implications (in section IV). The effort to formalize an analysis of apartheid in terms of a competitive interest group model of social choice should be of interest not only to students of South African, but also to economists more generally concerned with endogenous public policy. The South African case provides a laboratory for this type of analysis because the boundaries which delineate interest groups are very clearly drawn.


It has long been recognized by economists and scholars of South Africa that the apartheid system, representing a comprehensive set of regulations affecting all aspects of economic, social and political life, arose essentially as a response on the part of the white working class to the threat of black labor market competition.(2) Although apartheid itself is phenomenon of the post-1948 era of National Party rule, it is really only the most recent phase of a long history of white labor elitism and black exclusion, brought about through the medium of an interventionist, statist polity characterized by a racially limited franchise. The harnessing of the instruments of state power to further the interests of the white labor force, at the expense not only of blacks but also of white capitalists, dates from the earliest period of industrialization in South Africa, which followed the discovery of precious minerals in the 1860s and 1870s.

The first manifestation of a formal "color bar" occurred when white unskilled and semiskilled mineworkers, many of them imbued with the socialist and laborite traditions of their British trade union heritage, resisted efforts by mine owners to substitute black workers who were willing to work for wages considerably below the "civilized" reservation wage of the white.(3) At the same time, Afrikaner farmers were concerned about the depletion of the stock of rural blacks and the consequent rise in farmworkers' wages due to the magnetic attraction of the mines and secondary industries which sprang up in the growing urban centers concentrated on the Reef and in the Orange Free State mining areas.(4) It was precisely a coalition of convenience between these two groups--white labor and rural Afrikaner nationalist--that succeeded, through its numerical dominance of the white electorate, in attaining control of the reigns of government in 1924, and thenceforth proceeding to erect the extensive lattice of color bar legislation that characterized the 1920s and 1930s and provided the prototype for modern apartheid.

The political economy of South African race segregation therefore has much to do with the economics of deep-level gold mining on the one hand and Afrikaner semi-feudal tenant farming on the other hand. As long as labor costs comprised a significant share of total production costs in both of these industries, there would be pressure on the part of mine owners to displace expensive white workers with lower-wage blacks, and there would be pressure on the part of white farmers to retard the process of black urbanization (see Lipton [1985, 88, 116]). With the capture of the instruments of state power by white workers and farmers, the stage was set for massive public intervention in the workings of the labor market. This intervention took two forms: (i) geographical restrictions on black property rights (thus creating a permanent migrant labor force that oscillated continuously between residential domicile in reserves or "homelands" and employment in "white" urban areas), and (ii) industrial relations legislation, such as racial job reservation, "rate for the job" rules and union certification procedures, that protected white workers from being 'undercut" by blacks.(5)

Race discrimination, in South Africa as elsewhere, therefore possesses a distinctly "economic" (as opposed to irrational, atavistic or purely prejudicial) rationale, as pointed out by Hutt [1964, 27-31]. Race prejudice is often politicized as a tool to make transfers of wealth from the powerless to the powerful (see Sowell [1983, 246-47]). Prejudice and other barriers to entry into a group (such as skin color, language and religion) serve to reduce the costs to political entrepreneurs of organizing racial or ethnic coalitions and of monitoring the behavior of group members who may attempt to undermine its cohesion or dilute group solidarity.(6)

In South Africa, segregationist policies emerged in answer to the fears of white workers and farmers at a time of rapid industrialization. This occurred via the translation of the preferences of the median white voter into state policy. The result was a massive bureaucracy whose raison d'etre was the production of market regulation designed to effect wealth redistributions away from blacks and white industrial and commercial capitalists in favor of white workers and agricultural capitalists. The antagonism of capital owners (both foreign and indigenous South African) to segregationist policy has been well documented in the historical and theoretical literature. Regulation of the labor market by the South African state prevents the allocation of labor and other inputs to their highest valued uses therefore "taxes" the returns to capital (including human capital). The owners of capital have not surprisingly resisted apartheid in its modern guises as well as its early segregationist predecessors.(7) This fact refutes the radical and Marxist contention that apartheid is a particular mode of class exploitation developed by South African capitalists, and, as Hazlett [1987, 43] points out, it also invalidates the belief of many well meaning anti-apartheid activists outside of South Africa that the withdrawal of foreign capital from that country (disin vestment) will remove the prime progenitor of apartheid practices. In fact, economic theory and actual events demonstrate unequivocally that apartheid is anticapitalist.(8)

That the racial politics of South Africa are the immediate offspring of the configuration of economic interests of various pressure groups is attested to by the amazingly pragmatic and flexible application of segregationist policy over time.(9) Several scholars, including Hutt [1964, 85], Sowell [1981, 38-9], Olson [1982, 173], Lipton [1985, 3-4] and Kendall and Louw [1987, 91], have pointed to the close correlation between racial integration and the competitive pressures consequent upon economic growth. During periods of rapid growth in South Africa, the demand for semiskilled and skilled labor has typically exhausted the pool of white workers and has drawn large numbers of blacks, Indians and Coloreds (people of mixed race) into the relatively high-paying industrial workforce. In fact, it was precisely these dynamics of the labor market and the business cycle which ushered in the era of true apartheid in 1948. Boom conditions during and after World War II had brought with them a rapid modernization of the South African economy, with manufacturing industry stimulated by the natural protectionism afforded by wartime isolation of the country from traditional suppliers. The influx of an increasingly settled black force into the urban population centers now seemed to be an irreversible reality (see Lipton [1985, 21]). Either the white electorate would have to accept the inevitable deracialization and integration associated with capitalist development and market forces, or regulatory intervention by the state would be required to stem the tide (see Doxey [1961, 187-88]). The 1948 elections showed that the median white voter favored the latter course of action, and the new National Party government embarked on its ambitious project of central planning which Hazlett [1987, 16] describes as a sort of ethnic socialism. This comprised the establishment of "grand apartheid"--meaning rigid geographical separation (creation of separate sovereign national entities in which blacks would hold citizenship)--together with thoroughgoing regulation of economic and social interactions between blacks and whites in the "white" parts of the country where black workers supposedly sojourned temporarily.

But even this imposing edifice of bureaucratic statism has not been completely unconstrained by the costs associated with suppressing efficient market transactions and the general tide of economic growth. Continuous modernization and expansion of the industrial sector of the South African economy has increased the demand for highly skilled labor well beyond the capacity of the white population alone to supply it, so that some significant concessions have been made to the need for a stable, educated (and therefore characteristically middle class) black labor force. Some of the most severe apartheid laws have been retrenched since the late 1970s. Black unions are now accorded official recognition in the state industrial relations infrastructure, the pass laws which regulated the system of "influx control" have been repealed, racial miscegenation is no longer illegal, political parties are now permitted multiracial membership, blacks have been granted property rights in some "white" areas, and some business zoning restrictions affecting nonwhites are no longer in force. While the essential pillars of apartheid remain intact at time of writing (e.g., the Group Areas Act, which provides for racially segregated residential districts, and the Population Registration Act, which categorizes all citizens by race), the reforms that have occurred reflect the shifting structure of economic interests within the white electorate. Lipton [1985, 7] points out that labor costs are now a much smaller proportion of total production costs in an increasingly capital-intensive economy, so that there is less incentive for employers to replace white workers with cheaper black labor. This, in turn, has diminished the need for artificial protection of a beleaguered white working class, constantly threatened by low black wages. Furthermore, many whites, including Afrikaners, have been co-opted into the capitalist class, largely as a result of their acquisition of human capital. Owners of capital (human or physical) have no interest in regulations that inhibit the rational allocation of complementary labor inputs.

Thus, apartheid policy can be viewed along a continuum of different levels or degrees of application. Apartheid (and segregation more generally) has taken different forms, and has been enforced to differing extents, depending on economic and political circumstances. This suggests that apartheid policy is a pragmatic response of a rational white oligarchy which has faced different exogenous constraints over time. Hutt [1964, 30] argues convincingly that racial dogma and jingoistic white nationalism are not the underlying causes or motive forces driving apartheid--they are merely invoked when convenient to justify state interventions in the market economy that are intended to bring about material redistributions in favor of the politically dominant group. In the next section, a collective choice framework will be developed in order to formalize an interest group analysis of apartheid as endogenous policy.


The function of an interest group state is to specify and enforce property rights which maximize the wealth of politically influential groups, even if total societal wealth is diminished in the process.(10) Regulations are enacted by the state if it perceives that the marginal gain to beneficiary groups exceeds the marginal disutility to loser groups. The political equilibrium is therefore determined by the relative effectiveness of competing interest groups at producing political influence--i.e., controlling free riding.(11) It is nevertheless likely that the interest group state will design the institutions of wealth transfer in such a way as to achieve redistributions with minimum deadweight costs (see Becker [1983, 387]).

In constructing a competitive interest group framework with which to analyze the apartheid state, I assume (following Porter and Lundahl) that apartheid manifests itself in the South African-type economy in the form of government-enforced restrictions on the geographical and occupational mobility of black workers.(12) Black mobility is limited by a large number of regulations--including influx control, urban residential zoning restrictions, and industrial location policy--so that the cost of hiring black laborers is effectively increased. (Workers are not so much locked into a particular urban location as they are compelled to migrate back and forth between the rural reserve and the industrial town.) In 1986, influx control legislation (the so-called "pass laws" and the Black Urban Areas Consolidation Act) was repealed.(13) The mobility of black workers is still constrained, however. The pass laws were replaced by restrictions on urbanization ostensibly intended to control the spread of squatter slums.(14) Furthermore, as pointed out in section II above, two key pieces of legislation, which together facilitate many apartheid practices, remain in force--namely the Group Areas Act and the Population Registration Act.(15)

As a simplification, I assume that the chief pro-apartheid interest group comprises white unskilled and semiskilled workers and farmers. The chief anti-apartheid interest group within the white electorate comprises owners of physical and human capital. I also assume that vote-maximizing politicians attach influence weights to the preferences of these two groups which are proportional to the sizes of their memberships. The state then chooses optimal levels of the apartheid policy variables--namely, the level of enforcement of apartheid laws or the extent of implementation of apartheid policies--as well as the tax rates levied on workers and capital owners to finance apartheid, subject to the constraint that total tax revenue equal the cost of administering, maintaining and defending the apartheid system, which itself is inversely related to black utility. The more black workers express their disaffection by becoming active in the political resistance, the greater the costs to the white state of maintaining and policing the apartheid edifice.(16) Furthermore, an upward shift of this "apartheid cost function" might be precipitated by an exogenous shock that raises domestic and external opposition to apartheid, such as riots, strikes by black workers, or perhaps economic sanctions imposed by other nations (see Kaempfer and Lowenberg [1986; 1988a]).

The approach outlined above suggests that, ceteris paribus, any increase in the costs of defending or administering apartheid will result in a decrease in the optimal "level" of apartheid. However, an increase in the degree to which white workers benefit from apartheid, measured by an increase in the proportion of the white electorate comprising workers as opposed to capital owners, produces an increased "demand" for apartheid (enforcement of apartheid practices will become more rigorous).

Cyclical fluctuations of aggregate output in the South African economy are very sensitive to the world price of gold.(17) A rise in the gold price generates a positive output effect throughout the economy, which reduces the need on the part of white workers for costly public regulation protecting them from black labor market competition. In other words, the marginal utility of apartheid institutions to the median white worker decreases when he becomes wealthier, so that apartheid principles can be applied more leniently.

Thus, to summarize, the major exogenous variables which determine the level of enforcement of apartheid policies are: (i) parameters of the defense, policing and administrative costs of apartheid, (ii) the share of workers as opposed to capital owners in the white electorate, and (iii) the gold price.


The purpose of this section is to indicate how the specification of several time series regression equations might be a useful start toward testing some of the implications of the approach discussed in section III. The empirical analysis is not based on correctly-identified structural model, however. It is thus inevitable that relevant variables will be missing from equations, which precludes any causal inferences. Under such conditions, Desai [1968, 2-6] points out that little more can be offered than statements about "tendencies towards correlation." The best that can be hoped for is to provide what Mayer [1980, 173] refers to as "circumstantial evidence" in favor of the hypotheses in question.

The "level" of apartheid can be measured either by its results (e.g., the number of blacks employed in the industrial sector or the number of blacks receiving advanced training) or by the extent of enforcement of apartheid regulations (e.g., the application of "pass law" or influx control restrictions). The former measure is likely to be correlated with economy-wide cyclical conditions and sectoral shifts, as well as changes in social policy. For this reason, the latter method is used here, with variables measuring the degree of implementation or enforcement of apartheid laws acting as proxies for the level of apartheid. The sample period is restricted to the years during which classical" or traditional apartheid prevailed, and excludes the period after 1981 during which major structural changes in the apartheid system occurred (As mentioned above, the 1980s have been marked by several policy reforms of varying importance.)

Given the large number of possible explanatory variables which might be correlated with the level of apartheid, a stepwise estimation procedure is used. This makes it possible to screen out the less important explanatory variables and to concentrate attention on the major exogenous variables identified in section III. The regressions reported represent only a small sample out of a large number of alternative specifications; they are the end product of a process of pretesting or data mining which, as Mayer [1980, 174] indicates, is unavoidable when so little information as to correct model specification is available. The danger of data mining is that it is likely to lead to an overstatement of significance levels.(18) Therefore, t-statistics should be treated as no more than sample-descriptive, and any pretense of estimating structure and hypothesis testing should be avoided, as suggested by Lovell [1983, 10]. The purpose of this section is a modest one--namely, to show how the various dimensions and determinants of apartheid might be measured, thereby giving some operational meaning to the theoretical approach of section III. Regression results are presented.(19) A two-stage least squares procedure is used to identify two functional relationships, each of which is analogous to a "demand" for apartheid on the part of beneficiary interest groups. Use a similar method to estimate apartheid supply" or cost functions. It begins with two ordinary least squares regressions, which together comprise the first state of the two-stage process. The dependent variables are total real expenditure by South Afrrica government institutions in the black homelands (HOMELANDS COSTS) and real per capita defense expenditure (DEFENSE COSTS).(20) These are both intended to proxy the price" of apartheid to the white state. The first is an administrative cost associated with one of the key aspects of traditional "grand apartheid" (the homelands policy); and the second is the cost of defending the apartheid system in the face of black resistance, particularly in the form of guerrilla insurgencies.

The two main explanatory variables used are dummies for the Portuguese decolonization of neighboring countries in 1974 and the Soweto riots of 1976. I hypothesize that, when the Portuguese granted independence to Angola and Mozambique, the costs of defending white southern Africa from exiled resistance movements increased significantly. Similarly, the Soweto riots led to an awakening of domestic black opposition and thus increased the costs of enforcing and administering the system of apartheid laws. Both of these events therefore should have produced an upward shift of the apartheid cost function, in which case each of these two dummy variables should bear a positive relationship to both measures of the price of apartheid (HOMELANDS COSTS and DEFENSE COSTS). The results indicate that the Portuguese decolonization dummy and the Soweto dummy are indeed positively related to the two price variables, but only in the case of Portuguese decolonization is the relationship statistically significant. Both dummy variables are lagged by one year to reflect the assumption that politicians and bureaucrats do not adjust the in behavior instantaneously to changes in constraints. A linear time trend (TIME) and lagged values of the dependent variables are also included as regressors. (It should be noted that the DEFENSE COSTS regression displays negative serial correlation of the error terms, which leaves the coefficient estimates unbiased but overstates the t-statistics.)

The two regressions reported in part suggest that any event which raises the costs of administering, enforcing or defending the apartheid system will produce an upward shift of the apartheid cost function, or the supply" curve of apartheid, and an increase in the "price" of apartheid. This shift of the supply curve can then be used to identify the demand relationship between the price of apartheid and the quantity or level of apartheid enforcement. Part B of Table I contains a two-stage least squares regression in which the fitted values of HOMELANDS COSTS, as estimated,(21) are used to explain the level of apartheid measured by the variable PASSLAWS. The latter is the number of persons prosecuted for offences relating to pass law or influx control regulations, per 1,000 of the black population (The relevant items of legislation include the Black Urban Areas (Consolidation) Act of 1945 and the Black Labor Relations Regulation Act of 1953, or their various successors.)(22) PASSLAWS is thus intended to measure the degree of enforcement of apartheid, or, in some sense, the "quantity" of apartheid. Using the same two-stage least squares procedure to regress PASSLAWS on the fitted values of DEFENSE COSTS as estimated. in part A.

The demand curve for apartheid is indeed downward sloping. In part B there is statistically significant negative relationship between the quantity of apartheid (PASSLAWS) and the fitted values of the price of apartheid (HOMELANDS COSTS). Likewise, in part C there is a statistically significant negative relationship between the level of apartheid and the fitted price variable (DEFENSE COSTS).

Similar stepwise analysis is used to identif an upward sloping apartheid cost function. It starts with an ordinary least squares regression (the first step of this two-stage process) in which the dependent variable is BAN. The latter is defined as the number of "banning orders" in force against individuals. When the South African government "bans" or "restricts" people, it typically places them in a state approximating house arrest and excludes them from participating in political activity. Historically, banning orders have been enforced under the Suppression of Communism Act of 1953, the Riotous Assemblies Act of 1956, and the Internal Security Act of 1976.(23) I hypothesize that banning represents a significant enforcement and maintenance cost of the apartheid system, not only in terms of the resources used to implement banning orders through the judicial system and to monitor the behavior of banned persons, but also in terms of the loss of goodwill incurred domestically and internationally when this highly visible form of restriction is used. Furthermore, I expect that any increase in the level of apartheid requires that more resources be allocated to restricting political activity (banning), because higher levels of apartheid generate more resistance activities and necessitate additional restraining actions by the state. Thus the variable BAN can be treated as a measure of the supply price of apartheid.

The two main independent variables used in part A of Table II are the number of white full-time students enrolled in South African universities per 1,000 of the white population (WHITE EDUCATION), and the real gold price per ounce (GOLD PRICE). The analysis in sections II and III suggests that as white workers become more educated and acquire more human capital, they will demand less apartheid because they will no longer value protection from black labor market competition. In fact, educated whites might even lose from apartheid if human capital is complementary to black labor in production. Thus an increase in the rate of human capital accumulation by white (as measured here by university enrollment) should constitute a downward shift of the demand curve for apartheid policy. Similarly, section III suggests that an increase in the gold price, which makes whites (and blacks) better off, reduces the demand on the part of whites for state-imposed restrictions on the black labor force.

Essentially, I hypothesize that both an increase in the white university enrollment variable and an increase in the gold price result in a downward shift of the demand curve for apartheid, in the first case because of an occupational shift within the white electorate (a decrease in the number of whites who are workers as opposed to capital owners), and in the second case because of an economy-wide wealth effect. If the apartheid cost function has the expected shape, then a downward shift of the demand curve for apartheid implies a decrease in the supply price of apartheid as measured by the variable BAN. There is indeed a negative relationship between BAN on the one hand and both WHITE EDUCATION and GOLD PRICE on the other hand, although only in the former case is the relationship statistically significant. Again, this regression uses a linear time trend, one-year lags on the independent variables, and lagged values of the dependent variable.

In part B of Table II, the shift of the apartheid demand curve (induced by exogenous changes in the occupational structure of the white electorate and the gold price) is used to identify a positively sloped apartheid cost or supply function. Again, the variable PASSLAWS is used to measure the "quantity" or level of enforcement of apartheid. An alternative quantity measure is also introduced--namely LABOR BUREAUS, which is the number of Indians and Coloreds placed in employment through government labor bureaus per 1,000 of the Indian and Colored population. Labor bureaus were, until the abolition of the pass laws in 1986, one of the chief instruments of influx control, and the volume of their activity can therefore be regarded as a useful indicator of the extent to which apartheid policy was being enforced.(24) Both measures of the "quantity" of apartheid (PASSLAWS and LABOR BUREAUS) are regressed on the fitted values of the BAN variable, as estimated using a two-stage least squares procedure. Both quantity measures bear a statistically significant positive relationship to the supply price proxy, BAN, indicating an upward sloping apartheid cost function.

Taken together, Tables show that it is possible to measure at least some dimensions of the apartheid polity, and thereby test the proposition that the extent of application of race policy in South Africa is endogenous to changes in the parameters identified in the theoretical approach of section III.


The existence of apartheid is to be understood as an endogenous policy outcome emerging from the configuration of political influences exerted by the interest groups which comprise the white South African electorate. The extent to which apartheid institutions are enforced is a continuous dependent variable, suggesting that apartheid is not something which either exists or does not exist, but is rather a political commitment (a vector of policies) whose intensity has fluctuated over time. The magnitude of this variable is determined through the operation of a number of objective constraints perceived by the state, including the costs incurred in administering, defending and policing the apartheid system (which depends on the degree of domestic and external resistance). Other relevant exogenous variables are gold price and the occupational structure of the while electorate. The previous section has demonstrated that, while it is difficult to measure a concept as nebulous as "apartheid." the theoretical approach developed in section III is not without some operational empirical content.

The analysis has interesting implications for the survivability of the a partheid state(25) and the effectiveness of economic sanctions in achieving the dismantling of apartheid. In fact, Kaempfer and Lowenberg [1986; 1988a] use a similar method to analyze the effects of investment sanctions and an oil boycott on the South African economy and polity. If the apartheid system represents an essentially rational response of a heterogeneous white oligarchy to the costs and benefits associated with alternative institutional arrangements, then it follows that the system is likely to undergo pragmatic adjustments to changing constraints. There will be no tendency for the white ruling group to resist intransigently all such change, but it is also unlikely that apartheid restrictions will be liberalized in the absence of the necessary changes in exogenous variables. Recent increases in the level of black resistance within South Africa (which may be partly the result of increased foreign pressure against apartheid) have given rise to policy changes that would have been inconceivable before.(26) As Lipton points out, [1985,235-471], the costs of apartheid to the state have increased in the 1970s and 1980s.

The approach proposed in this paper has important implications for the debate on the desirability of imposing economic sanctions on South Africa.(27) The real purpose of the sanctions campaign might be to appease domestic interest groups in the sanctioning country seeking protectionist benefits or desirous of making moral statements, and might have very little to do with the actual impact of sanctions on the target country--as argued by Kaempfer, Lehman and Lowenberg [1987a; 1987b] and by Kaempfer and Lowenberg [1989]. Nevertheless, to the extent that the protagonists of sanctions hope to achieve political reform in the target country, it is necessary to examine the effects of sanctions on the political economy of that country. Careful study of the political economy of apartheid provides clear evidence of the fact that apartheid is essentially a comprehensive system of public regulation imposed on a market economy with the purpose of making redistributions in favor of a politically dominant subset of the white electorate at the expense of the wealth of the nation as a whole. The literature on the history of apartheid reveals that as long as the economy experiences healthy growth, the demand for regulatory protection of the white working class from black competition diminishes. Recession, whether induced by a cyclical downturn or by an external event such as sanctions, intensifies the demand for apartheid regulation on the part of white labor and farming interests. If an unhampered and flourishing market economy is a powerful catalyst for the breakdown of race discrimination, then any trade intervention (such as sanctions) which distorts resource allocation and inhibits the possibilities for voluntary exchange in factor and goods markets holds a serious risk of entrenching socially regressive policies. Nevertheless, even sanctions which are largely symbolic and do not inflict severe economic damage might precipitate desirable policy changes if they communicate signals to the opponents of apartheid which encourage increased political resistance and thus produce an upward shift of the apartheid cost function.

(*) Associate Professor of Economics, California State University, Northridges, I thank Charles Berker, William Brown, Adam Gifford, Ben Heijdra, Robert Krol, Eugenia Toma, Mark Toma, Alan William an Ben Yu for comments on earlier versions. The editor, Thomas Borcherding, and anonymous referees provided may helpful suggestions for improvement.

1. For example, see Porter [1978, 749, 754] and Lundahl [1982, 1178].

2. See Van der Horst [1942, 157], Doxey [1961, 114-15], Hutt [1964, 59], Lipton [1985, 187] and Kendall and Louw [1987]. Van der Horst, Doxey and Hutt represent classic treatments of the economic origins of political support for apartheid, while Lipton provides an invaluable modern account of how the structure of economic interest groups in South Africa has determined the evolution of race policy. Kendall and Louw demonstrate how the existing apartheid system is the product of far-reaching government regulation that is wealth reducing for society as a whole.

3. See Doxey [1961, 75, 116], Hutt [1964, 59-63] and Kendall and Louw [1987, 37-40].

4. See Van der Horst [1942, 295] and Lipton [1985, 108].

5. The process of institutionalization and legalization of the color bar and geographical separation is described by Doxey [1961, 91-5, 128,-57] and Hutt [1964, 72-8, 154].

6. This point is made by Borcherding [1977, 54-5]. See Roback [1987] for a most illuminating discussion of the use of state power as a means to create discriminatory practices which have public good attributes. In their analysis of the distributional effects of apartheid regulations, Knight and McGrath [1977, 271] show that race discrimination in South Africa exists precisely because it serves the interests of particular groups at the expense of others. It is important to note, however, that neither the black group nor the white group is homegenoues, and interests differ within groups as well as between them.

7. See Doxey [1961, 111], Sowell [1983, 112-13], Olson [1982, 163-64] and Lipton [1985, 112-16]. It should be noted that, until relatively recently, white capital owners werelargely English speaking (of British descent), which created another basis of antagonism between them and Afrikaner nationalists.

8. Hutt [1964, 30, 44] maintains that Afrikaner social psychology, rooted in Calvinist doctrine, is intrinsically opposed to the "capitalist spirit" because it is based on a fatalistic respect for historically determined race and class differences. Competitive models of discrimation in labor market reveal that total white income is always reduced by discriminatory practices, although white workers benefit at the expense of white capitalists. See Becker [1976, 19].

9. The pragmatism of apartheid policy is discussed in depth by Adam [1971].

10. On the concept of an interest group state see Stigler [1971], Peltzman [1976], North [1981, 22] and Olson [1982, 44].

11. See Kaempfer and Lowenberg [1988b].

12. Historical evidence for the realism of this assumption is provided by Doxey [1961, 133,163-64] and by Kendall and Louw [1987, 41-6].

13. See the Economics, April 26, 1986.

14. See the Daily News of Durban, April 23, 1986.

15. For details of the traditional legislative and administrative instruments of apartheid policy see Jones and Griffiths [1980], Steenkamp [1983], Giliomee and Schlemmer [1985], or Lowerberg [1984, 205-20].

16. This is based on Albert Breton's [1974, 86] notion of the "desire for redress" on the part of nonruling groups of citizens, which, to the extent that it translates into actual or potential opposition to the state, is percieved as a costs associated with a given policy package.

17. On the vital role played by the gold price see Kaempfer and Moffett [1988].

18. For example, Lovell [1983, 10-11] shows that a data miner who has no priori knowledge about which variables are the most important in explaining the phenomenon under study, and who preceeds to test at a nominal 95 percent significance level after picking the two explanatory variables with the highest t-coefficients out of twenty possible candidates, should realistically claim a 40 percent probability of committing a Type I error if the candidate explanatory variables are orthogonal. Exaggerated claims of significance are likely to be even more pronounced whenthe candidate variables are collinear.

19. Population data required to construct per capita series exclude the population of the Transkei after 1977 and of Bophutatswana after 1978. Gold price data are based on the average of daily fixings in the London market. The data are obtained from the following sources: South Afrrica, Bureau of Statistics [1965; 1966]; South Africa, Department of Information [1974-83]; South Africa, Department of Statistics [1968; 1978]; South African Institute of Race Relations [1958-82]; South African Reserve Bank [1961-82]; Savage [1977]; Mare [1980].

20. HOMELANDS COSTS is measured in thousands of constant (1953) rands, while the per capita variable DEFENSE COSTS is measured in constant (1953) rands.

21. The first state regreassion in Part A is reported for illustrative purposes only. In fact, the computer programs estimates both stages together and reports only the final two-stage least squares results. This two-stage procedure obtains fitted values of HOMELANDS COSTS from a reduced form equation in which the endogenous variable HOMELANDS COSTS is regressed on allexogenous variables--including HOMELANDS COSTS (-1) and PASSLAWS(-1). It is this fittedHOMELANDS COSTS which is used as an explanatory variables in the second stage regression.

22. See Jones and Griffiths [1980, 88-90].

23. I was unable to obtain observations of this variable for four years (1957, 1958, 1960 and 1961), and these values are linearly interpolated.

24. See Jones and Griffiths [1980]. It is possible, however, that this variables might reflect business cycle conditions as well as apartheid enforcement.

25. See Johnson [1977], Gann and Duignan [1978] and Hanf, Weiland and Vierdag [1981].

26. See Kaempfer and Lowenberg [1986, 390-91].

27. For recent treatment of the effects of sanctions on the South African economy and apartment policy see C. Becker [1987], Hazlett [1987] and Kaemfer and Lowenberg [1986; 1988a].

REFERENCES Adam, Heribert. Modernizing Racial Domination. Berkeley and Los Angeles: University of

California Press, 1971. Becker, Charles M. "Economic Sanctions Against South Africa." World politics,January 1987,

147-73. Becker, Gary S. The Economic Approach to Human Behavior. Chicago: University of Chicago

Press, 1976. ------."A Theory of Competition among Pressure Groups for Political Influence." Quarterly

Journal of Economics, August 1983, 371-400. Borcherding, Thomas E. "An Economic Approach to School Integration: Public Choice with

Tie-ins." Public Choice, Fall 1977, 53-77. Breton, Albert. The Economic Theory of Representative Government. Chicago: Aldine Publishing

Co., 1974. Daily News."Urbanisation: New Strategy," Durban, South Africa, April 23, 1986. Desai, Meghnad. "Some Issues in Econometric History." Economic History Review, April 1968,

1-16. Doxey, G. V. The Industrial Colour Bar in South Africa. Cape Town: Oxford University Press,

1961. Economist. "Apartheid Isn't Dead," London, April 26, 1986, 15. Gann, Lewis H. and Peter Duignan. South Africa: War, Revolution or Peace? Stanford: Hoover

Institution Press, 1978. Giliomee, Hermann and Lawrence Schlemmer. "The Influx Control Fence," in Up Against the

Fences: Poverty, Passes and Privilege in South Africa, edited by Hermann Giliomee and

Lawrence Schlemmer. Cape Town: David Philip, 1985. Hanf, Theodor, Heribert Weiland, and Gerda Vierdag. South Africa: The Prospects of Peaceful

Change. London: Rex Collings, 1981. Hazlett, Thomas W. "American Sanctions on South African Apartheid." Program in Applied Public

Policy Research Working Paper No. 1, University of California, Davis, Institute of

Governmental Affairs, January 1987. Hutt, W. H. The Economics of the Colour Bar. London: Andre Deutsch Ltd., 1964. Johnson, R. W. How Long Will South Africa Survive? New York: Oxford Universitty Press, 1977. Johnston, J. Econometric Methods (Third Edition). New York: McGraw-Hill, 1984. Jones, Robert A. and Howard R. Griffiths. Labour Legislation in South Africa . Johannesburg:

McGraw-Hill, 1980. Kaempfer, William H., James A. Lehman, and Anton D. Lowenberg. "Divestment, Investment

Sanctions, and Disinvestment: An Evaluation of Anti-Apartheid Policy Instruments."

International Organization, Summer 1987a, 457-73. ------."The Economics of the Call for Anti-Apartheid Investment Sanctions." Social Science

Quarterly, September 1987b, 528-38. Kaempfer, William H. and Anton D. Lowenberg. "A Model of the Political Economy of

International Investment Sanctions: The Case of South Africa." Kyklos 39(3), 1986, 377-96. ------."Sanctioning South Africa: The Politics Behind the Policies." Cato Journal, forthcoming,

Winter 1989. ------."South Africa's Vulnerability to Oil Sanctions." Working paper, Claremont Center for

Economic Policy Studies, 1988a. ------."The Theory of International Economic Sanctions: A Public Choice Approach." American

Economic Review, September 1988b, 786-93. Kaempfer, William H. and Michael H. Moffett. "The Impact of Anti-Apartheid Sanctions on

South Africa: Some Trade and Financial Evidence." Contemporary Policy Issues, October

1988, 118-29. Kendall, Frances and Leon Louw. After Apartheid: The Solution for South Africa. San Francisco:

Institute for Contemporary Studies, 1987. Knight, John B., and Michael D. McGrath. "An Analysis of Racial Wage Discrimination in South

Africa. "Bulletin of the Oxford University Institute of Economics and Statistics, November

1977, 245-71. Lipton, Merle. Capitalism and Apartheid: South Africa, 1910-84. Totowa, NJ: Rowman and

Allanheld, 1985. Lovell, Michael C. "Data Mining." Review of Economics and Statistics, February 1983, 1-12. Lowenberg, Anton D. "Towards an Economic Theory of the Apartheid State." Ph.DD. dissertation,

Simon Fraser University, 1984. Lundahl, Mats. "The Rationale of Apartheid." American Economic Review, December 1982,

1169-79. Mare, Gerry. African Population Relocation in South Africa. Johannesburg: South African Institute

of Race Relations, 1980. Mayer, Thomas, "Economics as a Hard Science: Realistic Goal or Wishful Thin king? "Economic

Inquiry, April 1980, 165-78. Nattrass, Jill. The South African Economy: Its Growth and Change. Cape Town: Oxford University

Press, 1981. North, Douglass C. Structure and Change in Economic History. New York: W.W. Norton, 1981. Olson, Mancur. The Rise and Decline of Nations: Economic Growth, Stagflation, and Social

Rigidities. New Haven, CT: Yale University Press, 1982. Peltzman, Sam. "Toward a More General Theory of Regulation." Journal of Law and Economics,

August 1976, 211-40. Porter, Richard C. "A Model of the South African-Type Economy." American Economic Review,

December 1978, 743-55. Roback, Jennifer. "Racism as a Collective Good (or Bad)." Manuscript, Center for Study of Public

Choice, George Mason University, January 1987. Savage, Michael. "Costs of Enforcing Apartheid and Problems of Change." African Affairs, July

1977, 287-302. South Africa, Bureau of Statistics. Statistical Yearbook. Pretoria: Bureau of Statistics, 1965, 19 South Africa, Department of Information. Official Yearbook of the Republic of South Africa.

Pretoria: Department of Information, 1974-1983. South Africa, Department of Statistics. South Africa Statistics. Pretoria: Government Printer,

1968, 1978. South African Institute of Race Relations. Survey of Race Relations in South Africa. Johannesburg:

South Africa Institute of Race Relations, 1958-1982. South African Reserve Bank. Quarterly Bulletin. Pretoria: South African Reserve Bank,

1961-1982. Sowell, Thomas. The Economics and Politics of Race: An International Perspective. New York:

William Morrow and Company, 1983. ------."Markets and Minorities. New York: Basic Books, 1981. Steenkamp, W. F. J. "Labour Problems and Policies of Half a Century." South Arican Journal

of Economics, March 1983, 58-87. Stigler, George. "The Economic Theory of Regulation." Bell Journal, Spring 1971, 3-21. Van der Horst, Sheila T. Native Labour in South Africa. London: Frank Cass and Company, 1942.
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:Lowenberg, Anton D.
Publication:Economic Inquiry
Date:Jan 1, 1989
Previous Article:A test of steady-state government debt neutrality.
Next Article:Bank reserve requirements as an impediment to signaling.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters