Despite several weaknesses, the profile of social protection that emerges from this analysis supports general observations and points to possible strategies for action.(Endnote 22) The observations all relate to the enormity of the challenges facing the new government in extending social protection -- the financial crisis looming over the state pension scheme, the heavy dependence of the South African workforce on lump-sum retirement benefits, and the existence of a massive population of illegal migrant workers. Given their broad scope, these problems cannot be expected to disappear from the horizon anytime soon, and the constraints that they impose on improving social protection are severe.
At the same time, there are compelling reasons for revising the social protection that South Africa extends to its migrant population. The challenge is to identify approaches which are affordable and distribute costs equitably. Both the reasons for change and some possibilities for new policy directions are described below.
a. The case for change
The most compelling reasons for changing the two-gate system have to do with equity. Three improvements could be achieved.
First, change could bring social protection for migrants into closer conformity with both the spirit of and letter of South Africa's new democracy, which generally calls for equal rights and benefits for all residents. Significantly, only a limited number of rights bestowed by the new Constitution extend to citizens only, in particular those relating to participation in political decision making and governance. Among the specific guarantees in the Constitution is a right to "health care, food, water, and social security" [italics added].(Endnote 23) Extending to all residents, this provision provides a strong legal basis for changing the two-gate system.
Second, change could achieve greater horizontal equity among migrants. As shown in section II, the legal distinction between temporary and permanent workers has been severely blurred in practice. Tens of thousands of "temporary" migrants have been working in South Africa for years, and some, for decades. Given this situation, there is no clear rationale for denying them social protection while extending it to workers who entered under the Aliens Control Act. A similar argument can be made for migrants who entered South Africa or chose to remain there illegally prior to 1991. Many of these individuals now have strong roots in South African communities, including long-term employment, spouses, children, and even grandchildren. To deny them social protection today based on their violation of a law whose error has been recognized and repealed would be to place them in double jeopardy.(Endnote 24)
Third, change could serve to "level the playing field" between migrants and South African workers. Because social protection financed by employer contributions can increase the cost of labour, its absence or near absence is one of the factors that can lead employers to favour migrants in hiring.(Endnote 25) While South Africa's limited reliance on employer contributions (only for the UIF and Workers Compensation fund) means that this cost differential is small, benefit improvements that are now under consideration will widen the gap. The National Labour and Economic Development Institute (NALEDI) has argued,
The best way of preventing undercutting and exploitation is to ensure that all workers earn the appropriate minimum wage for that sector, work the same hours, and receive the same fringe benefits [italics added]. Extending the regulatory net makes it less attractive for employers to employ cheap foreign labour.(Endnote 26)
While these observations focus on privately-provided benefits, their logic applies with equal force to social protection.
b. Possible directions
A major barrier to the realization of these benefits is an economic disincentive that confronts governments in the region, discouraging them from extending protection to migrants in the absence of parallel action by other governments. Known as a Prisoner's Dilemma, this disincentive exists because each government can realize the maximum benefit for itself and its citizens (including those who migrate) if it declines to extend social protection to migrants while neighboring governments do so. This outcome is optimal from each individual government's perspective because it minimizes its own social protection expenditures while enabling it to benefit from others' expenditures for its migrant workers. Yet if every government withholds social protection on the expectation that others will provide it, migrants will be essentially unprotected, disadvantaging all parties. In this situation, the most promising approach to extending benefits is government-to-government agreements which bind all parties to coordinated action.
(1) Benefit improvements through governmental agreements
(i) Reciprocity - The ILO Conventions on social protection of migrants take reciprocity as their starting point.(Endnote 27) They provide that protection is to be granted only to migrants from states which themselves, by ratifying the pertinent Conventions, agree to extend protection to migrants from other countries. In other words, a state which refuses to extend protection to another state's migrants cannot expect that state to grant protection to its workers. As for the scope of benefits, the Conventions generally call for equal treatment, meaning that social protection for migrants should be "as similar as possible" to that which the country provides for its citizens.
Useful in some contexts, these notions have limited applicability in Southern Africa at present. Three barriers stand in the way of their use. First, social protection in most neighboring countries is still relatively undeveloped. This limits their governments' ability to offer significant concessions at the bargaining table and, as a consequence, makes them unlikely to win much in the way of reciprocal concessions. Second, South Africa's reliance on the private sector to provide benefits on a discretionary basis limits its own capacity to negotiate concessions for migrants as a group. Third, the Conventions generally apply to contributory social insurance schemes. As a means-tested payment funded by general revenues, South Africa's retirement benefit, the state pension, falls outside their scope.
(ii) Political linkages - Limited social protection does not, however, preclude the negotiation of mutual concessions among governments; it rather requires greater innovation in identifying mutual interests. There is, for example, considerable latitude for governments to link different types of benefits in devising agreements or to link social protection with other issues. For example, countries that have contributory pension schemes -- Mozambique and Zimbabwe at present, soon to be followed by Namibia and Zambia -- can still use these to negotiate with South Africa on the benefits it does provide. With increasing numbers of South African farmers migrating to Mozambique, the Mozambican government might extend old age and disability coverage to them in return for South Africa's extension of improved retrenchment benefits to returning Mozambican migrants. Under a more diverse linking of concessions, South Africa might extend some form of retrenchment assistance to migrants in return for their home country's loosening of local content and labour requirements for South African firms working on site there.
(iii) Regional integration - Integration of social protection can help eliminate the trap described in the previous section [II(c)(3)], ineligibility at home because of work abroad. This is particularly important because proposals for freer movement of citizens, if adopted, would cause larger numbers of workers to face this problem. The ILO Conventions provide three rules for integration. Together these ensure that no migrant is excluded from protection
because his or her work history spans national boundaries and, conversely, that no worker receives duplicative protection for this reason. The rules are:
* determination of applicable legislation, which requires specification of which country's law governs the social protection of migrants in cases of ambiguity or conflict;
* maintenance of acquired rights and provision of benefits abroad, which means that any current right or earned prospective right should be guaranteed to the migrant in either country, even if it has been acquired in the other; and
* maintenance of rights in course of acquisition, which means that where a right is conditional upon the completion of a qualifying period, account should be taken of periods served by the migrant worker in each country.
One means of promoting integration would be the establishment of a board or subcommittee under SADC's auspices. This group could both encourage ratification and provide a forum for resolving practical issues of compliance. In addition, this group could suggest common program definitions, application procedures, and waiting periods, thereby encouraging greater conformity among schemes.(Endnote 28) A natural starting point for this effort would be workers compensation, since it exists in some form throughout the region. An issue for early consideration would be eliminating double coverage for migrants working for a home firm on site in a neighboring country, possibly through the establishment of a region-wide agreement under which they could opt for coverage either at home or in their country of assignment.
(2) Benefit improvements through unilateral government action
As social protection develops in Southern Africa, cooperative approaches such as reciprocity, linkages, and regional integration will become more useful in extending benefits to migrants. In the meantime, is there a rationale basis for unilateral action by South Africa to improve migrant benefits? Though this question requires some speculation, the preceding analysis suggests three such possibilities.
(i) Shaping social protection for migrants to address South African objectives - It may be possible to achieve incremental benefit improvements if they are structured to produce advantages for both the migrant population and South African citizens. Recognizing these mutual benefits, South Africa might find it advantageous to act unilaterally, thereby overcoming the Prisoner's Dilemma. A possibility suggested in the previous section [II(c)(1)] is the structuring of social protection to help combat the underlying causes of migration -- unemployment, economic stagnation, underdevelopment, and lack of training in neighboring countries. Taking this approach, South Africa could not only assist migrants but also reduce the political and fiscal stress it is experiencing as the result of a sharply expanding migrant population.
The efficacy of development efforts as a tool for controlling migration is recognized in the South African Reconstruction and Development Plan (RDP). It holds that:
In the long run, sustainable reconstruction and development in South Africa requires sustainable reconstruction and development in southern Africa as a whole. Otherwise, the region will face continued high unemployment and underemployment, leading to labour migration and drain to more industrialized areas.(Endnote 29)
This recognition also exists in other quarters. The Labour Market Commission endorsed the notion of regional development integration on the grounds that it would reduce migration by increasing output and productivity in the region as a whole. Two labour umbrella organizations, NUM and COSATU, have called for stemming migration through a Regional Reconstruction and Development Program (RRDP) aimed at combating unemployment and economic stagnation in neighboring countries.
Structured to promote development, retrenchment benefits might be provided to migrants as a combination of (1) a lump-sum cash payment to help cover the costs of resettlement, and (2) subsidies for job creation in the migrant's home country. Both forms of support could be financed by extending the UIF levy (in 1997, one percent of wages up to R76,752) to migrant workers and their employers.(Endnote 30) The proceeds from the levy could be accumulated in a separate trust within the UIF. Existing UIF offices could take applications for resettlement payments and determine migrants' eligibility. The payments could then be remitted to qualified migrants who returned home through local TEBA offices throughout the region. Subsidies for job creation might be paid to migrants' home governments or to NGOs. A model for such subsidies exists is South Africa's support for Mozambican "agri-farming" projects [section II(c)(1)], which provide returning migrants with farm equipment and seeds. Combined with short-term cash assistance, such projects could effectively bridge the gap between social protection and regional development, ensuring that the allocated funds not only help to replace lost wages but also increase employment, output, and purchasing power in the economy to which the migrant returns(Endnote 31).
(ii) Improving implementation of existing schemes - In addition to experiencing problems related to ineligibility for benefits, migrants also have difficulties resulting from flawed implementation of social protection which South African law provides them. These problems relate both to the two-gate system [section II(a)] and to the payment of benefits as remittances [II(c)(4)]. To better fulfill the promises of existing law, the South African government could take three actions:
* Base eligibility for social protection on the duration of a migrant's stay - The present, two-gate entry system could be replaced with a more even-handed means of establishing eligibility for social protection. In its place, the government might require that all migrants, whatever their formal classification, begin participating in contributory social protection after they have been working in South Africa for a set time period.(Endnote 32) This approach would treat all workers identically, ensuring that social protection is not denied long-term migrants because they bear the formal label of "temporary." It would also be in keeping with the Labour Market Commission's broader proposals which call for complete elimination of temporary contract labour as a legal category.
* Require regular audits of Workers Compensation remittances - As the single social protection scheme which covers all migrants, workers compensation is in need of measures to increase transparency and reduce the fractured character of administration. To increase transparency, the government could enact a statutory requirement for annual audits of all parties, public and private, charged with transmitting compensation payments to migrants and their families on its behalf. Such audits should focus on random samples of claims, as Rand Mutual did in investigating remittances to Mozambican migrants [section II(c)(4)]. An audit requirement should also be coupled with a mandate to disseminate the results broadly -- to Parliament, to migrants' home governments, and publications with broad readerships in all the affected countries.
* Establish a migrants' ombudsman in the Labour Department - Finally, the South African government could both improve its own service to migrants and that of their labour representatives by establishing an ombudsman's office within the Department of Labour. This office would be broadly changed with liaison with the migrant population. It would respond to requests for information, investigate complaints, and facilitate timely eligibility decisions and benefit payments. It would also provide space, training, and administrative support to the labour representatives. By moving these representatives a step
closer to the main administrative agency with which they must deal, these measures would both strengthen their capacity and provide the department with more direct feedback on its performance.
(iii) Contributory financing as a prerequisite to benefit improvements
Defined in section I, social protection schemes are in essence mechanisms for sharing risk based on the principle of social solidarity. The reality in most countries is that social solidarity is stronger among citizens than between citizens and migrants. As a result, the host country's willingness to subsidize migrants' social protection is limited. This limitation implies that migrants as a group must cover a significant portion of the cost of the benefits they receive. Financing which relies on worker and employer contributions is essential to achieving this outcome.(Endnote 33)
Yet, as shown in section I, a striking feature of social protection in South Africa is limited reliance on contributory financing. While the UIF is funded by a one percent payroll withholding from workers, means-tested state pensions for retired and disabled workers are financed entirely by general tax revenues. As a result, middle and high income taxpayers subsidize the pensions paid to low income individuals.
Given this financing configuration, the potential for a workable extension of social protection to migrants differs by program. It would be comparatively simple to structure retrenchment benefits as an extension of the UIF since its financing is contributory and administrative mechanisms are in place to collect revenues and pay benefits. Establishing a contributory scheme for providing old age and disability protection to migrants would be more difficult at present, since there is no national pension scheme on which to build.
Of course this is not the only problem associated with the government's exclusive reliance on general revenues to finance means-tested state pension benefits. As shown in section I, coverage is uneven; many workers deplete their assets to qualify for benefits; and, as a consequence, program costs are increasing sharply. Together these problems are fueling increased interest in some form of contributory financing for old age and disability protection. The need for improved social protection for migrants constitutes yet one more reason to move in this direction.
Endnote 22:
The weaknesses are, first, that it relies heavily on existing sources of information. Additional research could cast light on many questions on which there is little or no existing data, such as the ways in which illegal migrants cope, or fail to, in the absence of social protection. Second, the private, optional status of employment-related retirement and disability benefits in South Africa means that coverage is uneven, both across the South African workforce and the migrant population. This unevenness makes it difficult to arrive at generalizations. Third, the absence of any national statistics on the social protection of migrants further confounds efforts to generalize. Finally, the limited information which is available relates heavily to a single industry, mining, where private provision of benefits is relatively extensive. Focusing what exists may thus convey an excessively rosy impression of the benefits that migrants receive.
Endnote 23.
Constitution of the Republic of South Africa, 6 May 1996, section 27(1).
Endnote 24:
As noted in section II, the South African government has already taken two steps in this direction.
However, both are limited in that the time periods for coming forward are short and the group that can qualify
for permanent resident status is tightly restricted.
Endnote 25:
Toolo, Hilton and Lael Bethlehem (researchers at the National Labour and Economic Development Institute),
"Migration to South Africa: problems, issues, and possible approaches for organized labour," Centre for Policy
Studies, Workshop proceedings no 14, 10 April 1995.
Endnote 26:
Toolo, Hilton and Lael Bethlehem, NALEDI, p. 34.
Endnote 27:
These are Conventions 19, 48, 118, 157, and 165.
Endnote 28:
Its focus would be on standard administration rather than equalizing benefit levels, which will probably
continue to vary throughout the region reflecting countries' differing fiscal capacities.
Endnote 29:
Reconstruction and Development Plan, in Dolan, Chris, "Policy Challenges in New South Africa," in
Southern African Migration Domestic and Regional Implications, Centre for Policy Studies, Workshop
proceedings no. 14, page 57.
Endnote 30:
This extension would reduce the cost advantage to employers of hiring migrant workers, thus helping to
"level the playing field" between migrants and nonmigrants in South African labour markets.
Endnote 31:
Non-contributory schemes -- i.e., the state pension -- call for a different approach, to be dealt with subsequently.
Endnote 32:
A major consideration with fiscal implications for this proposal is the so-called "GEAR" (The Growth
Employment and Redistributive Strategy) and " MTEF" (The Medium Term Expenditure Framework i.e., multi-year budgeting cycle). GEAR sets targets for the key macro-economic indicators - The budget (government
spending), interest rates, investment levels, inflation, employment, and other economic factors. therefore, any
potential increase in expenditure will be capped for the following three years. Except to allow for inflation. This
is likely to pose fiscal limits on migration policy.
Endnote 33:
In the United States, for example, migrants as a group pay more in social security contributions than they
receive in benefits, providing a net subsidy for the Social Security Trust Funds. This results from the
combination of (i) a near universal requirement for withholding of contributions from payrolls, and (ii) illegal
migrants' reluctance to come forward to collect benefits.
For further information, please contact the ILO Office at Pretoria at Tel: +27.12.341.2170, Fax: +27.12.341.2159 or by E-mail: pretoria@ilo.org
Copyright © 1998 International Labour Organization (ILO)
Disclaimer
webinfo@ilo.org
This page was created by RS. It was approved by FL. It was last updated on 21 April 1998.