To the Editor:
Thomas Sowell’s provocative article, “‘Affirmative Action’: A Worldwide Disaster” [December 1989], points to many weaknesses in the theory and practice of affirmative action. But since there is no given definition of affirmative action, it is not a difficult exercise to locate unsound ideologies and disastrous programs under that broad umbrella. Nor is it difficult to identify good policies and positive outcomes. . . . Does the existence of ideological faults, weak implementation, and distortion justify the sweeping conclusions suggested by the title? Surely it does not. The article does not face up to positive results or to affirmative-action objectives based on individual rights. . . .
1. A scheme to minimize handicaps owing to group membership may involve group-based preferences. This does not require the achievement of statistical parity, which may be neither feasible nor equitable, since the disparity may be partly due to cultural differences. On the other hand, gross imbalances might suggest the probability of significant group-based disabilities and the possibility of some discrimination. Such discrimination may be due to a variety of factors which may or may not involve prejudice on the part of the decision-maker (as analyzed by Kenneth Arrow, Thomas Schelling, Alan Wertheimer, and others). But the justification for affirmative action is based not on the apportionment of blame but rather on the existence of group-based advantages and disadvantages. The main objective is not to compensate for wrongs suffered but to help overcome certain group disabilities such as those related to ethnicity and gender.
2. Mr. Sowell questions the validity of claims for historical compensation. This factor has been urged by some as imparting a moral imperative to affirmative action. Are U.S. blacks entitled to reparations from U.S whites because their current handicaps in relation to whites are partly attributable to past oppression of blacks by whites? Is Israel . . . entitled to reparations from the countries of Europe for the past oppression of Jews in those countries? There are many interesting studies on the subject, . . . but the case for affirmative action does not rest on acceptance of the principle of reparations.
3. Many of Mr. Sowell’s explanations for the persistence of great intergroup inequalities establish rather than undermine the case for affirmative action. Major interethnic imbalances in one generation may lead not only to significant inequality in life chances for the next generation but also to prejudice and discrimination. Even if there were initially no hostility or prejudice, it would be rational to use ethnicity as a screening device in selection. . . . Thus the implementation of President John F. Kennedy’s Executive Order to “take affirmative action to ensure that the applicants are employed, and that employees are treated during employment without regard to race, creed, color, or national origin” may require active group-specific intervention irrespective of the existence of any prejudice on the part of the employer.
4. Mr. Sowell’s defense against charges of racial discrimination in employment is based on the fact that there are large differences in educational opportunities: e.g., if blacks and whites go to different kinds of colleges and take up different disciplines, their qualifications will indeed have different values and employers will be justified in recognizing such differences. But this clinches the case for affirmative action to minimize the handicaps suffered on account of ethnicity even more surely than the charge of discrimination. Likewise, Mr. Sowell’s defense against charges of caste discrimination in admission to Indian medical schools is based on the existence of severe inter-caste imbalances in educational facilities from infancy. This would suggest that the cure should be made even more potent. Perhaps there should be a series of interventions starting in primary school rather than a single intervention at the point of entry into medical school.
5. There are many kinds of inequality in all societies. It may be argued that it is desirable to reduce all of these and not just the ones based on ethnicity and gender. But different kinds of policies are needed to tackle different kinds of inequalities. A program to reduce one type of inequality may aggravate another. For example, an indiscriminate scheme of “merit-based” scholarships in secondary schools may open avenues of mobility for certain sections of the urban lower-middle class and significantly reduce one type of imbalance in opportunity; but the net impact on certain other imbalances (interethnic or pertaining to gender or urban-rural differences) may be negligible or even negative.
A sophisticated program (or a package of programs) may need to be designed depending on the weights we assign to bridging the different types of gaps in opportunity. Extra weights need to be assigned to narrowing those disparities which tend to be self-perpetuating, e.g., through creating prejudices on account of ethnicity or gender. In turn, the remedies, whether or not they serve other socially desirable goals, are unacceptable if they are based on, or are likely to contribute to, prejudice or inequality.
In lumping together apartheid policies designed to reinforce a racial hierarchy in South Africa and affirmative-action policies designed to eliminate hierarchies in other societies, Mr. Sowell disregards the fundamental distinction between legitimate and illegitimate preferences which Ronald Dworkin, for example, has analyzed in his book, Taking Rights Seriously.
6. Mr. Sowell suggests that all improvements followed trends that were in place prior to affirmative action. There is indeed evidence that many of the changes designed to be achieved through preferential policies occurred before the enforcement machinery came into operation, but this is not a phenomenon peculiar to affirmative action. In the U.S., as in many other countries, there is a substantial willingness to follow leads given by national movements, by the political leadership, by the legislature and higher judiciary. Many institutions quickly and voluntarily detect and adapt to new trends which may be set prior to and irrespective of any enforcement. Enforcement mechanisms may consolidate or reinforce a trend or, by heavy-handed action, provoke a backlash, but the origins of the trend are often to be found elsewhere.
In the U.S., the civil-rights movement of the third quarter of this century, the Kennedy and Johnson presidencies, radical federal legislation, and a remarkable sequence of landmark judgments by the Supreme Court set a new trend and created an atmosphere which evoked a strong positive response from other political leaders, the state legislatures, the lower courts, public-sector agencies, private firms, educational institutions, professional bodies, and the public. That the implementation of affirmative action ran ahead of the legislation and enforcement in the environment of the 1960’s does not imply that affirmative-action policies made no difference, but rather that there was wide acceptance of them in that decade.
7. Like many other policies, affirmative action has had both successes and failures in conception and implementation. Mr. Sowell attributes the failures to affirmative action and all progress to other factors. He suggests that preferential policies have undermined the integrity of Pakistan, for instance. But prejudice and discrimination against East Pakistanis were so open and widespread that the more credible conclusion would be that it is the failure to implement affirmative action which has contributed to Pakistan’s troubles, including the secession of Bangladesh. He goes on to suggest that the riots against backward castes in Bombay are indicative of the failure of affirmative action. They are indeed linked to affirmative action but may be manifestations of the efforts by the high castes to stop what they recognize as progress toward achieving its objectives. Are the campus riots in the U.S. the consequences of affirmative action, or are they caused by the failure to implement affirmative action in the 80’s and the related resurgence of racism?
8. Mr. Sowell is surely right in suggesting that disturbing ethnicity-based hierarchies and attempting to reduce interethnic gaps may increase friction. Black-white relations in the U.S. were smoothest under slavery, and even after Emancipation, there was less racial tension when Jim Crow laws prescribing segregation and regulating activity were mutually accepted.
In encouraging new aspirations, in promoting dissatisfaction with institutional obstacles to mobility, and in undermining the legitimacy of segregation and the established hierarchy, civil-rights ideologies undoubtedly contributed to provoking the riots which accompanied Brown and many of the civil-rights advances that followed. Jews, too, faced hostile, often violent, reaction when they sought to overcome barriers to their mobility in the early decades of this century. In India and South Africa also, every attempt to break down caste barriers or race barriers has invariably generated friction, which has frequently sparked off violence. How can it be otherwise? It is, of course, desirable to reduce tension and avoid riots, but whether or not we should do so by curtailing reforms depends on what kind of society we want and what price we are willing to pay to achieve that objective.
9. In the U.S. and elsewhere, affirmative action has had successes and failures. The picture is complex and varied, as shown in the works of scholars such as Marc Galanter in his monumental study of affirmative action in India, Competing Equalities. In sum, Mr. Sowell’s claim that affirmative action “has been a failure in the U.S. and a disaster in other countries” which have had such policies longer is unwarranted by the evidence.
Devanesan Nesiah
Ministry of State for Rehabilitation and Reconstruction
Colombo, Sri Lanka
_____________
Thomas Sowell writes:
Devanesan Nesiah’s argument begs all the questions and, above all, the central question: what do we know and how do we know it? His claim that my article has “no given definition of affirmative action” blithely ignores the definition given in the very first paragraph: “government apportionment of coveted positions, to supersede the competition of the marketplace or of academia.” It is he who has no definition of the crucial terms of his argument, such as “imbalance,” “disabilities,” or “group-based advantages and disadvantages.” Operationally, do “disadvantages,” for example, mean institutional barriers ex ante or statistical results ex post? In Mr. Nesiah’s own country, Sri Lanka, it was the Tamils who were “disadvantaged” ex ante but the Sinhalese who were “disadvantaged” ex post. After preferential policies, both were devastated in the bloody riots that escalated into civil war.
More specifically, let me answer him point by point.
- To say that statistical differences between groups “might suggest” something nebulously called “disabilities” in addition to “the possibility of discrimination” is interesting speculation for a seminar. But are courts of law supposed to award multimillion-dollar damages on the basis of speculation—especially since those possibilities are offset by Mr. Nesiah’s admission that statistical differences “may be partly due to cultural differences”? Demographic and locational differences are also among many other factors behind statistical differences. Are courts supposed to be playgrounds for theories or places where serious claims and serious consequences depend on evidence and proof?
- Mr. Nesiah says nothing on the issue of “historical compensation” to justify my going through the arguments again. The readers of COMMENTARY are intelligent enough not to need a repetition—and not to be thrown off the trail by Mr. Nesiah’s red herring.
- Once again, Mr. Nesiah’s speculation about what “may” happen slides into an assumption about what does happen and what policy should be.
- Speculation turns to fiction when Mr. Nesiah discusses “Mr. Sowell’s defense against charges of racial discrimination.” There is no such defense, either in this article or in anything else I have ever written. The question Mr. Nesiah never faces is: what is to be regarded as evidence in a court of law, in a case involving specific litigants? Moreover, if he thought seriously about what he said when he referred to “handicaps suffered on account of ethnicity” and how that “clinches the case for affirmative action,” then consistency would require him (and others) to conclude that Jews, Chinese, and Japanese should also be beneficiaries of quotas, instead of being victimized by them. But this assumes that phrases like “handicaps suffered on account of ethnicity” are meant seriously, not simply used as political rhetoric. Do “handicaps suffered on account of ethnicity” include internal cultural handicaps as well as external institutional discrimination? Transient as well as permanent handicaps? Nor does the undefined “intervention” tell us much about what ought to be done, or why it would be expected to make matters better rather than worse.
- The distinction Mr. Nesiah makes between the intentions of some preferential policies and others has no relevance to my study of the effects of such policies. The similarities in effects are all the more striking because of the enormous differences in intentions, countries, and groups involved.
- Mr. Nesiah’s desperate efforts to evade the plain fact that trends widely attributed to affirmative action in fact antedate it once more subsitute speculation for evidence. In the case of blacks, the rise in income and occupations antedated affirmative action by decades. Surely, Mr. Nesiah is not suggesting that people were rushing to get on a bandwagon that had not yet been designed, much less built. Unfortunately, this clinging to social dogmas in the face of devastating facts to the contrary is not confined to Mr. Nesiah. The National Academy of Science’s widely acclaimed book, A Common Destiny, says: “In terms of per-capita incomes, family incomes, and male workers’ earnings, blacks gained relative to whites fairly steadily from 1939 to 1969.” Then came a change: “Since the early 1970’s, the economic status of blacks relative to whites has, on average, stagnated or deteriorated.” Ironically, these data and dates show more racial economic progress before affirmative action than afterward. Yet this book reaches all the conventional conclusions, including a favorable view of affirmative action.
- I have nowhere said that affirmative action “undermined the integrity of Pakistan.” Almost equally reckless is Mr. Nesiah’s suggestion that it is “the failure to implement affirmative action” in American colleges and universities which is responsible for a “resurgence of racism” on campus. Statistics on racial incidents show that they are most prevalent at liberal and radical institutions where there is more racial preference. Massachusetts alone had more such incidents than the entire South. Moreover, in many places the current levels of racial hostility and violence exceed anything seen in past generations on those campuses—however much Mr. Nesiah’s dogma requires him to see it as a “resurgence.”
- If closing the gap between whites and minorities were the cause of friction and violence, then it is Asian Americans who would be the principal targets of campus backlash, for they have not only closed the gap but exceeded the performance of whites. On some elite campuses, they have taken more places from whites than have blacks and Hispanics put together. Once more, Mr. Nesiah has not found it necessary to test his hypothesis against any evidence. The evidence here suggests that the old song was right: “It ain’t what you do; it’s the way that you do it.”
- Mr. Nesiah’s assertion that my conclusions were “unwarranted by the evidence” was only his second use of the word evidence. But he offered no evidence either time.
What is even more incredible than Mr. Nesiah’s convoluted reasoning to justify preferential policies is that his own country—which was once justifiably held up as a model for harmonious intergroup relations—was so polarized by preferential policies that its people were burning each other alive in the streets and the whole country ultimately degenerated into a civil war whose end is not yet in sight.