The late Michael Wallerstein models the distributive consequences of affirmative action:
The motivation for redistribution along racial lines in the model is an inefficiency in the labor market creating rents for those holding good jobs, coupled with an exogenously (historically) given difference in the distributions of human capital across races. When racial divisions lead to demands for redistributive policies along racial lines via affirmative action, we show that legislative policy bargaining implies that the amount of redistribution along income lines is less on average than would exist were racial divisions absent. When affirmative action is on the agenda, redistribution along racial lines partly replaces redistribution along income lines and total redistribution, as measured by the change of the Gini coefficient, declines.
The paper shows how a ruling elite might use affirmative action to implement a divide-and-rule strategy. It notes the strategic aim behind Richard Nixon’s approval of the Philadelphia Plan and creation of the Office of Minority Business Enterprise: ‘Nixon’s strategy introduced a new policy dimension that split black workers and white union-members, establishing a Federal program of race-based hiring [and procurement] that has never been revoked.’
Wallerstein mentions the rent-sharing aspect of affirmative action, without elaborating much on this point. I want to do so in this post.
In the bargaining game of affirmative action, the parties (state on one side and interest group on the other) come to a rent-sharing arrangement to divide the surplus arising from some economic activity. Here the state represents a diverse coalition of the proprietary classes, which for strategic reasons (consolidating political stability and securing economic advantage, as Wallerstein shows) allows the other party (elite and professional middle-class members of some disadvantaged group) to capture a share of booty.
The rents may be created by monopoly privileges arising from government economic activity, or they may already exist due to segmented labour markets (e.g. in which, as Wallerstein points out for African Americans, members of one ethnic group are excluded from the labour market for secure, well-paid jobs with promotional ladders). They may also stem from artificially preserved shortages, such as professional guilds (e.g. medical associations) which allow members to earn scarcity premiums. Or rents may simply be generated by the ordinary competitive operations of private firms. An interest group, made up of elites driven by a broad sense of injustice alongside narrow careerist aspirations, uses ethnicity or sectarian identity to stake a claim to a share of these spoils.
The prize usually involves civil-service jobs and preferential admission to universities (which serve as passports to employment in the state bureaucracy, or to the professional salariat). It may also include business favours such as a dedicated share of government contracts, cheap loans, and allocation of licenses to operate exclusively in a particular market.
The US policy – giving what Nixon described as a ‘piece of the action’ to African Americans – has its equivalents elsewhere. In India, public-sector jobs are reserved and university quotas are granted to members of ‘backward castes.’ Malaysia’s New Economic Policy affords a varied set of privileges, including higher-education entry and bureaucratic positions as well as business permits and procurement contracts, to bumiputera (‘sons of the soil’). This privileging of Malay people and Malay-owned firms, at the expense of those with Chinese and South Asian ancestry, has an Indonesian equivalent in preferences for pribumi. Comparable Australian programs, granting discounted credit to businesses owned by people of Aboriginal or Torres Strait Islander ancestry, and seeking to win government contracts for Indigenous-owned businesses (‘supplier diversity… increases the opportunity for ethnic minority owned businesses to supply their goods and services to large public and private sector organisations’), are very much smaller.
On the other hand, Australia, like several other countries, has recognized a new form of property right based on the customary law of indigenous peoples. These vest communal title to land and water in corporate entities, enabling a small number of trustees or agents to exercise control rights over the assets. The previous post discussed a similar set of property rights in ‘ancestral land’ assigned to chiefs and heads of lineages (according to the ‘customary’ provisions of groups deemed to be indigenous) in various sub-Saharan African countries. Such institutional arrangements allow a few individuals to hold assets and exercise managerial powers (negotiation of land-use agreements and transactions with outside parties, investment, etc.) on behalf of beneficiaries to whom they are only weakly accountable. Usually these rights and obligations aren’t directly lucrative, as they don’t confer upon their bearer any entitlement to a share of the net income generated by the assets. Nonetheless they can be leveraged to obtain a stream of money and other benefits from consulting, directorships, government advisory panels, corporate liaison, etc.
These various group-specific privileges allow a tiny subset of the disadvantaged ethnic, linguistic or religious group to claim a portion of the social surplus extracted from other workers. Employment in a government agency usually involves incentive payments to motivate honesty: remuneration is set far above the incumbent’s reservation wage or salary, so that the cost of losing one’s job due to detected corruption is positive (i.e. exceeds the value of the public official’s next-best alternative). Even with mechanisms in place to discourage corruption, employment as a civil servant may be something of a sinecure, involving weak accountability and little public scrutiny. It thus presents opportunities to abuse public office to obtain private benefits through liaison and collusion with private actors. The government official may interact with business interests through the design and implementation of regulation, allocation of permits, subsidies and grants, dispensation of contracts for procurement, monitoring and inspection, etc. This allows bureaucrats to engage in rent-sharing, dividing with firms the spoils of patronage, graft and self-dealing, and establishing personal contacts that may allow a future stream of benefits. Even without explicit corruption, or the abuse of public office for private gain, a public monopoly generates rents which the government employee shares in. Public-sector jobs usually offer more job security and seniority benefits (guaranteed wage premiums and promotional ladders) than private-sector equivalents, privileges which may be especially important in countries without state-funded retirement income.
Similar benefits accrue from preferential access to university courses which allow entry to the liberal professions (i.e. in which graduates become certified academics, architects, lawyers, accountants and similar qualified practitioners). High training costs, or other barriers to entry, create a relative scarcity of these skills and enable their bearers to command a high income. Members of these professions generally work as independent contractors, often in small proprietorships or partnerships. As such they aren’t subject to the employment relationship (in which, in exchange for receipt of a wage or salary, the employee must submit to the authority, monitoring and direction of a manager, supervisor, or agent of the firm’s owners). Along with being residual claimants on the net income generated by their own activities, these professionals retain work autonomy and decison-making control of their production process, e.g. routines, effort, intensity, etc.
Thus the perceived over- or under-representation of Jews, Tamils, Igbo, diasporic Chinese, etc. in government jobs, among salaried professionals or in business interests becomes the stuff of tribalism and inter-group conflict. Do rent-sharing arrangements reduce envy and discontent and encourage ethnic (or linguistic, religious, gender, etc.) harmony?
In a bargaining framework, it’s easy to see that a ‘minority’-based interest group seeking to negotiate a rent-sharing agreement has an incentive to foment ethnic discord and encourage animosity. The bargaining power of each party, and thus the division of spoils resulting from negotiations, depends on the level of its reward (‘disagreement payoff‘) under alternative (non-cooperative) arrangements. A party advances its distributional claim (i.e. increases its bargaining power) if it can increase the costs or disadvantage inflicted on the other party that would result from a breakdown in bargaining and an implementation of the noncooperative outcome (‘threat point’). Political discord that threatens the stability of the social order and the favourability of the investment climate reduces the state’s (and propertied classes’) fallback position, and thus decreases the latter’s bargaining power. As shown by several of the country examples above, the prospect of a disadvantaged group (especially one whose demographic weight, economic position or geographic distribution makes it crucial to social stability) becoming radicalized may allow elite members of that group to extract favours from the ruling authorities. This is one reason why so many careerists have flirted with socialism, or declared a commitment to left-wing radical politics (only to emerge, a few years later, as a trusted member of the establishment, having passed through a nationalist phase). In return for these benefits, the elite in these groups preserves the prevailing order by channeling frustration along ethnic lines.
A credible threat of secession is a way to signal an interest group’s possession of a good ‘outside option’ (the reservation payoff it would receive, not just from ceasing to bargain, but from terminating the interaction entirely). In some circumstances, creation of a new state or political entity, especially in a resource-rich region (e.g. Katanga), would reliably grant civil-service jobs and commercial opportunities to elite members of the disadvantaged separatist group. In such cases the potential secessionists will be likely to win concessions and privileges to forestall or discourage their withdrawal from the original state. Of course, interactions are usually asymmetric in a quite different way: in most cases the ruling elite has the better outside option, and can make take-it-or-leave-it offers of very paltry prizes which the rent-seekers have little choice but to accept.
Indeed, in several of our examples rent-sharing arises from non-cooperative interactions rather than through bargaining. What this means is that the behaviour of at least one of the parties isn’t subject to binding agreements. In particular, any pledge by elite members of the disadvantaged or backward community to support status-quo institutions can’t be enforced. In such cases rents may be conferred unilaterally by one party (i.e. by the state and propertied classes) on the other, as a kind of incentive device. This regularly involves the cultivation of suitable people over decades, the identification of ‘trustworthy’ or reliable representatives at a young age, and their recruitment into a kind of internal labour market with its own job ladders and career trajectory. Such ‘community representatives’ then serve as bulwarks for the existing institutional setup, as it reproduces the income disparities, ethnic divisions and group-specific poverty which they purport to struggle against.