The high cost of affirmative action

IN JUNE 1998, PARLIAMENT is expected to vote on the Employment Equity bill. Once law, it will require firms and other organisations to start changing the composition of their workforces to achieve equitable representation of black people, women and the disabled. At a press conference last month, the Labour minister, Tito Mboweni, described the bill as one of the most important and central instruments to bring about change in the South African workplace. Much of the criticism directed at the measure, he said, had been based on ignorance rather than on an objective study.

That there are huge inequalities in the labour market is indisputable. The African unemployment rate for 1994 reached 41 per cent compared to rates of 6.4 per cent, 23 per cent, and 17 per cent for whites, coloureds, and Asians respectively. Furthermore, women fare worse in all racial groups, but African women are worst off with an unemployment rate of 50 per cent. These figures are reflected in incomes. The World Bank’s 1995 World Development Report showed that the poorest 20 per cent of South Africans received only 3 per cent of national income while 63 per cent went to the richest 20 per cent.

Supporters of affirmative action argue that the legislation will produce numerous other benefits besides eradicating inequalities of race and gender — improved industrial relations, greater respect for different cultures and values, less stress at work as employees fulfill their potential, increased product-ivity, more equitable income distribution leading to higher domestic demand, and greater economic growth. What the enthusiasts have not discussed, however, are the costs of achieving these desirable goals through affirmative-action law.

Fortunately, countries such as America and Malaysia have lengthy experience of affirmative-action law in pursuit of racial and gender equality — experience that we can draw upon to assess the costs and benefits that South Africans might expect from its proposals.
The United States passed its first affirmative-action law in 1964. Thirty years later, the outcomes, at best, are mixed.

In the unemployment arena the position of American blacks has actually got worse: in 1964 the ratio of black to white unemployment was 2:1, in 1990 it stood at 2.76:1. Nor has the economic status of minorities and women improved significantly. It is true that women’s share of professional degrees grew from 2.7 per cent in 1960 to 36 per cent in 1990 and that their average earnings as a percentage of men’s increased from 61 per cent to 72 per cent over the same period. However, a study of 138 firms by Alison Konrad and Frank Linnehan found that although companies with race-conscious policies had at least one woman at a higher rank, and more minorities in management positions than companies with race-neutral employment policies, in four other categories (the rank of the highest-ranking minority; the percentage of women in management; the percentage of women in the company as a whole; and the percentage of minorities in the company as a whole) the results were the same for both types of company.

Why did aggressively pursued affirmative-action policies produce such meagre gains? In his book The Affirmative Action Fraud (Cato Institute, Washington 1996) Clint Bolick, president of the Institute of Justice, explains that rather than helping new entrants to the lab-our market to find jobs, these measures largely shifted exist-ing employees from some employers to others.

The broader effects of the legis-lation also seem to have been limited. In a 1993 article in Forbes Magazine, Peter Brimelow and Lesley Spencer reported that the gap in black and white college participation rates has grown since the seventies. In 1976 about 22.6 per cent of blacks aged 18 to 24 were at colleges compared to 27.1 per cent of whites in the same age group; by 1990 black participation rates had grown to 25.4 per cent, while the rate for whites stood at 32.5 per cent.

Social conditions for blacks have also deteriorated. In 1950 only 9 per cent of black families were headed by a single parent; in 1965 that figure had increased to 28 per cent, and by 1993 to almost 50 per cent. In 1959 only 15 per cent of black births were illegitimate, in 1992, 66 per cent were. One in four black men in his twenties is either in jail, on probation, or on parole.

In terms of earnings distribution, the black middle class prospered during the period 1970-1990: the percentage of black families earning more than $50,000 per annum increased from 10 per cent to nearly 15 per cent. But this improvement might have happened without affirmative action, since the percentage of white families in high-income categories also grew during this period. Meanwhile, the percentage of black families earning below $15,000 in 1993 was nearly 40 per cent.

In Malaysia, affirmative-action policies to reduce the inequalities between the impoverished Malay majority and the ethnic Chinese minority have been in force for 20 years and can claim reasonable success. Progress has been made in establishing a Malay business community in an economy dominated by Chinese. And, according to a report by Ian Emsley published in the journal Optima, poverty decreased dramatically from 49 per cent of the population in 1969 to 19 per cent in 1987. Access to education and employment for Malays has also improved so that the distribution of employment in different sectors of the economy as well as in better-paying occupations is approaching target figures. But before South Africans start quoting these figures as proof of the effectiveness of affirmative action, it is essential to remember that the Malaysian economy was growing at a remarkable 6-7 per cent a year during these decades. This growth created new job opportunities to absorb the Bumiputra (“sons of the soil”) into the labour force while non-Malays did not lose their jobs to make room for Malays.

However, even after 20 years of affirmative action, Malays are not proportionally represented in management positions and the professions, and they still find themselves over-represented in the less productive parts of the economy. Ownership of equity has not reached the 1970 targets and although great strides have been made towards equalisation of income per capita between ethnic groups, income inequality within ethnic groups has widened. Richer Malays found it easier to use the opportunities in tertiary education, better-paid managerial positions, cheap shares, loans and government contracts provided by ethnic preferment. Once again, the beneficiaries of affirmative action have been predominantly those who already had a comparative advantage over their poorer compatriots. This is particularly pertinent for South Africa, where inequality within racial groups has grown steadily more significant and in 1991 accounted for 77 per cent of overall inequality.

These are some of the successes of affirmative-action policies in two countries, but at what cost were they garnered? Calculating their cost is notoriously difficult and reliable estimates are hard to come by. Easiest to assess are the direct costs: the bureaucratic expansion required to implement the affirmative-action measures. In America this machinery com-prises the Equal Employment Oppor-tunity Commission, the Office of Federal Contract Compliance, the Department of Education, and several other federal, state, and local agencies. Brimelow and Spencer found that the cost of regulation by these agencies amounted to approximately $545 million. When the costs of complying with affirmative-action regulations (including private-sector costs) are added, they estimate that direct costs rise to $17billion-$20 billion.

In South Africa the proposed affirmative-action institutional framework comprises the Department of Labour, the Directorate for Equal Opportunities, the Labour Inspectorate, industry-wide collective-bargaining forums, the Commission for Conciliation, Mediation and Arbitration, the Labour Court and the Labour Appeal Court, and the Employment Equity Advisory Council. Neither the Labour Market Commission report nor the Employment Equity green paper of 1996 contain any cost estimates for running this operation. In reply to a parliamentary question last month, Tito Mboweni said it would cost about R28 million a year to administer the bill, of which R15 million would be spent on inspections, R7 million on moni-toring, R3 million for the Employment Equity Commission and R3 million for policy development.

Besides direct costs, indirect costs and opportunity costs must be taken into account. Indirect costs involve no direct cash outlay but constitute real costs in the sense that delays in hiring and recruiting are caused by time-consuming affirmative-action regulations in, for example, the private sector and educational institutions. Such problems are already being experienced in the South African construction sector, where uncertainty about government’s affirmative-action requirements has resulted in long delays in allocating housing projects and a slowdown in growth in the building industry. Brimelow and Spencer estimate the indirect costs of US affirmative action for 1991 to be $96 billion.
Further indirect costs arise from interference with the market mechanism. In Malaysia, Emsley suggests that the allocation of human and financial resources “away from areas of highest return reduced possible productivity gains”.

He further observes that while inefficiencies arising from giving preference to and subsidising Malay firms may have been reduced by the practice of having non-Malay business partners and sub-contracting to Chinese firms, such behaviour did not encourage the development of entre-preneurship among Malays. In education, race-based preferential policies, particularly in awarding certificates, have led to the devaluation of certain degrees and the further complication that more prosperous Chinese have their children educated at international institutions. The result has been a double standard of education for Malays and non-Malays.

Brimelow and Spencer believe that in America the highest costs of affirmative action are the oppor-tunity costs — the opportunity cost of a particular choice is the value of the best alternative foregone — and estimate these at $236 billion for 1991. The opportunity costs of affirmative action arise from bad hiring decisions under government coercion, negative effects on morale, and the misallocation of financial resources. What results might have been achieved if these resources had been invested in efficient education, essential infrastructure or other efforts to empower the truly disadvantaged — those outside the economic mainstream?

Finally, there are the social costs, which are difficult to quantify but must not be overlooked. Affirmative-action policies based on race must use racial attributes to be implemented. This reinforces negative stereotypes, racial tension and a stigmatisation that thwarts the efforts of members of the preferred groups to pursue their goals on merit and hard work rather than preferential treatment.

The American affirmative-action landscape is littered with stories of abuse of the affirmative-action provisions, simply because race has become a proxy for disadvantage. In his book, No More Martyrs Now (Conrad Books), Don Caldwell cites the example of Harvard University’s financial aid to all minority students irrespective of need. When one of the beneficiaries was asked whether he had to maintain a certain standard of performance to receive the financial aid of the programme, he responded: “No, I have to prove I am still black!”

Race-based affirmative-action policies encourage a culture of entitlement that undermines initiative, self-confidence and self-reliance. The beneficiaries of racial preferences may always have to do more to prove their worth than others because of lingering suspicions of undue advantage. Even the beneficiaries themselves often wonder whether they are in demand because of their abilities or because they happen to be the “right” colour.
Shelby Steele, a black professor of English at San Jose State University in California, abhors the message that current affirmative-action policies send to young American blacks: “that extra entitlements are their due and that the greatest power of all is the power that comes to them as victims. If they want to get anywhere in American life, they had better wear their victimisation on their sleeve and tap into white guilt, making whites want to escape by offering money, status, racial preferences, anything — in return. Is this the way for a race that has been oppressed to come into their own? Is this the way to achieve independence?”

Should an argument against race-based affirmative action therefore be construed as an argument for the status quo? Certainly not. Efforts should be made to find ways in which all South Africans would become better off, but especially the truly disadvantaged — those with the least access to employment, food, security, and productive assets. Since the poorest of the poor are found predominantly in the rural areas and are more often than not women and youths, a sustained effort to help the rural poor to help themselves could go a long way towards solving problems of marginalisation and inequality. To a large extent, the success of the New Economic Policy in Malaysia can be credited to its commitment to rural development and particularly to efforts to raise the incomes of rural Malays through gains in agricultural productivity.

All South Africans must have access to rewarding employment, sufficient education and skills acquisition: these are non-negotiable requirements. But we cannot allow different standards based on race. Building capacity through hard work and patience will take longer, but it will certainly be a better use of resources than the attempted short-cut of affirmative action

Focus 10, April 1998. Rachel Jafta is lecturer in economics at the University of Stellenbosch. The complete text of this article is published in "Ironic Victory: Liberalism in post-liberation South Africa" edited by David Welsh and R.W. Johnson (Oxford University Press).